model_id,benchmark_id,score,reference_url,reported_setting_json,matches_canonical,source_type,audit_status,notes
claude-haiku-4.5,tau2_bench_airline,63.6,https://www.anthropic.com/news/claude-haiku-4-5,"{""effort"":""128K thinking budget"",""harness"":""\u03c4\u00b2-bench official"",""judge"":""official harness"",""mode"":""thinking"",""prompt_style"":""official + prompt addendum to Airline Agent Policy and User prompt (per Anthropic blog)"",""sampling"":""pass@1 (avg 10 runs)"",""temperature"":""default"",""tools"":""agentic (\u03c4\u00b2-bench official tools + tool use)""}",False,official_blog,verified,"Anthropic Haiku 4.5 blog (cross-model comparison table), airline domain, 128K thinking, avg 10 runs. Prompt addendum added to Agent Policy and User prompt for airline domain → matches_canonical=false."
claude-haiku-4.5,tau2_bench_retail,83.2,https://www.anthropic.com/news/claude-haiku-4-5,"{""effort"":""128K thinking budget"",""harness"":""\u03c4\u00b2-bench official"",""judge"":""official harness"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 10 runs)"",""temperature"":""default"",""tools"":""agentic (\u03c4\u00b2-bench official tools + tool use)""}",True,official_blog,verified,"Anthropic Haiku 4.5 blog (cross-model comparison table), retail domain, 128K thinking, avg 10 runs."
claude-haiku-4.5,tau2_bench_telecom,83.0,https://www.anthropic.com/news/claude-haiku-4-5,"{""effort"":""128K thinking budget"",""harness"":""\u03c4\u00b2-bench official"",""judge"":""official harness"",""mode"":""thinking"",""prompt_style"":""official + prompt addendum to Telecom Agent Policy and User prompt (per Anthropic blog)"",""sampling"":""pass@1 (avg 10 runs)"",""temperature"":""default"",""tools"":""agentic (\u03c4\u00b2-bench official tools + tool use)""}",False,official_blog,verified,"Anthropic Haiku 4.5 blog (cross-model comparison table), telecom domain, 128K thinking, avg 10 runs. Prompt addendum added to Agent Policy and User prompt for telecom domain → matches_canonical=false."
claude-sonnet-4,tau2_bench_airline,63.0,https://www.anthropic.com/news/claude-haiku-4-5,"{""effort"":""128K thinking budget"",""harness"":""\u03c4\u00b2-bench official"",""judge"":""official harness"",""mode"":""thinking"",""prompt_style"":""official + prompt addendum to Airline Agent Policy and User prompt (per Anthropic blog)"",""sampling"":""pass@1 (avg 10 runs)"",""temperature"":""default"",""tools"":""agentic (\u03c4\u00b2-bench official tools + tool use)""}",False,official_blog,verified,"Anthropic Haiku 4.5 blog (cross-model comparison table), airline domain, 128K thinking, avg 10 runs. Prompt addendum added to Agent Policy and User prompt for airline domain → matches_canonical=false."
claude-sonnet-4,tau2_bench_retail,83.8,https://www.anthropic.com/news/claude-haiku-4-5,"{""effort"":""128K thinking budget"",""harness"":""\u03c4\u00b2-bench official"",""judge"":""official harness"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 10 runs)"",""temperature"":""default"",""tools"":""agentic (\u03c4\u00b2-bench official tools + tool use)""}",True,official_blog,verified,"Anthropic Haiku 4.5 blog (cross-model comparison table), retail domain, 128K thinking, avg 10 runs."
claude-sonnet-4,tau2_bench_telecom,49.6,https://www.anthropic.com/news/claude-haiku-4-5,"{""effort"":""128K thinking budget"",""harness"":""\u03c4\u00b2-bench official"",""judge"":""official harness"",""mode"":""thinking"",""prompt_style"":""official + prompt addendum to Telecom Agent Policy and User prompt (per Anthropic blog)"",""sampling"":""pass@1 (avg 10 runs)"",""temperature"":""default"",""tools"":""agentic (\u03c4\u00b2-bench official tools + tool use)""}",False,official_blog,verified,"Anthropic Haiku 4.5 blog (cross-model comparison table), telecom domain, 128K thinking, avg 10 runs. Prompt addendum added to Agent Policy and User prompt for telecom domain → matches_canonical=false."
claude-sonnet-4.5,tau2_bench_airline,70.0,https://www.anthropic.com/news/claude-haiku-4-5,"{""effort"":""128K thinking budget"",""harness"":""\u03c4\u00b2-bench official"",""judge"":""official harness"",""mode"":""thinking"",""prompt_style"":""official + prompt addendum to Airline Agent Policy and User prompt (per Anthropic blog)"",""sampling"":""pass@1 (avg 10 runs)"",""temperature"":""default"",""tools"":""agentic (\u03c4\u00b2-bench official tools + tool use)""}",False,official_blog,verified,"Anthropic Haiku 4.5 blog (cross-model comparison table), airline domain, 128K thinking, avg 10 runs. Prompt addendum added to Agent Policy and User prompt for airline domain → matches_canonical=false."
claude-sonnet-4.5,tau2_bench_retail,86.2,https://www.anthropic.com/news/claude-haiku-4-5,"{""effort"":""128K thinking budget"",""harness"":""\u03c4\u00b2-bench official"",""judge"":""official harness"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 10 runs)"",""temperature"":""default"",""tools"":""agentic (\u03c4\u00b2-bench official tools + tool use)""}",True,official_blog,verified,"Anthropic Haiku 4.5 blog (cross-model comparison table), retail domain, 128K thinking, avg 10 runs."
claude-sonnet-4.5,tau2_bench_telecom,98.0,https://www.anthropic.com/news/claude-haiku-4-5,"{""effort"":""128K thinking budget"",""harness"":""\u03c4\u00b2-bench official"",""judge"":""official harness"",""mode"":""thinking"",""prompt_style"":""official + prompt addendum to Telecom Agent Policy and User prompt (per Anthropic blog)"",""sampling"":""pass@1 (avg 10 runs)"",""temperature"":""default"",""tools"":""agentic (\u03c4\u00b2-bench official tools + tool use)""}",False,official_blog,verified,"Anthropic Haiku 4.5 blog (cross-model comparison table), telecom domain, 128K thinking, avg 10 runs. Prompt addendum added to Agent Policy and User prompt for telecom domain → matches_canonical=false."
gpt-5,tau2_bench_airline,62.6,https://www.anthropic.com/news/claude-haiku-4-5,"{""effort"":""128K thinking budget"",""harness"":""\u03c4\u00b2-bench official"",""judge"":""official harness"",""mode"":""thinking"",""prompt_style"":""official + prompt addendum to Airline Agent Policy and User prompt (per Anthropic blog)"",""sampling"":""pass@1 (avg 10 runs)"",""temperature"":""default"",""tools"":""agentic (\u03c4\u00b2-bench official tools + tool use)""}",False,third_party,verified,"Anthropic Haiku 4.5 blog (cross-model comparison table), airline domain, 128K thinking, avg 10 runs. Prompt addendum added to Agent Policy and User prompt for airline domain → matches_canonical=false. Reported by Anthropic, not OpenAI primary — source_type=tech_report (third-party measurement on competitor model)."
gpt-5,tau2_bench_retail,81.1,https://www.anthropic.com/news/claude-haiku-4-5,"{""effort"":""128K thinking budget"",""harness"":""\u03c4\u00b2-bench official"",""judge"":""official harness"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 10 runs)"",""temperature"":""default"",""tools"":""agentic (\u03c4\u00b2-bench official tools + tool use)""}",False,third_party,verified,"Anthropic Haiku 4.5 blog (cross-model comparison table), retail domain, 128K thinking, avg 10 runs. Reported by Anthropic, not OpenAI primary — source_type=tech_report (third-party measurement on competitor model)."
gpt-5,tau2_bench_telecom,96.7,https://www.anthropic.com/news/claude-haiku-4-5,"{""effort"":""128K thinking budget"",""harness"":""\u03c4\u00b2-bench official"",""judge"":""official harness"",""mode"":""thinking"",""prompt_style"":""official + prompt addendum to Telecom Agent Policy and User prompt (per Anthropic blog)"",""sampling"":""pass@1 (avg 10 runs)"",""temperature"":""default"",""tools"":""agentic (\u03c4\u00b2-bench official tools + tool use)""}",False,third_party,verified,"Anthropic Haiku 4.5 blog (cross-model comparison table), telecom domain, 128K thinking, avg 10 runs. Prompt addendum added to Agent Policy and User prompt for telecom domain → matches_canonical=false. Reported by Anthropic, not OpenAI primary — source_type=tech_report (third-party measurement on competitor model)."
o3-mini-high,aime_2024,87.3,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o3-mini-high=87.3.
o3-mini-high,gpqa_diamond,77.2,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o3-mini-high=77.2.
o3-mini-high,codeforces_rating,2130.0,https://arxiv.org/abs/2504.21318,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4-reasoning paper Table 1 (o3-mini-high column): codeforces_rating=2130 (re-sourced from https://arxiv.org/abs/2504.21318, matches BP value)"
o3-mini-high,mmlu,86.9,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o3-mini-high=86.9.
o3-mini-high,math_500,97.9,https://github.com/openai/simple-evals,"""MATH-500, 0-shot (footnote: MATH-500 not full MATH for newer models)""",True,tech_report,verified,
o3-mini-high,humaneval,97.6,https://github.com/openai/simple-evals,"""HumanEval, 0-shot""",True,tech_report,verified,
o3-mini-high,simpleqa,13.8,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: o3-mini-high=13.8.
o3-mini-high,hle,13.0,https://scale.com/leaderboard/humanitys_last_exam,{},,,dropped," [Dropped: Model not found in current Scale.com HLE leaderboard (50 entries, 2026-04-29 scrape). Ghost cell per R5g-a-ghost-cell.]"
o3-mini-high,frontiermath,12.4,https://epoch.ai/benchmarks/frontiermath,"{""effort"":""high"",""mode"":""thinking"",""tools"":""none""}",True,leaderboard,verified,epoch.ai FrontierMath-2025-02-28-Private: o3-mini-2025-01-31_high = 12.4% (mean_score=0.124). Canonical effort=high matches.
o3-mini-high,aime_2025,86.5,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: o3-mini-high=86.5.
o3-mini-high,chatbot_arena_elo,1363.0,https://lmarena.ai/leaderboard/text,{},True,leaderboard,verified,lmarena.ai text arena (fetched 2026-04-29). Arena model name: o3-mini-high (rank 144). ELO updates continuously; score reflects latest available.
o3-mini-high,arena_hard,81.9,https://arxiv.org/abs/2504.21318,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,Phi-4-reasoning paper Table 2 (o3-mini-high column): arena_hard=81.9. [R5d: prior unverified value 43.0 from https://github.com/lmarena/arena-hard-auto deleted.]
o3-mini-high,hmmt_feb_2025,67.5,https://arxiv.org/abs/2504.21318,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,Phi-4-reasoning paper Table 1 (o3-mini-high column): hmmt_2025=67.5. [R5d: prior unverified value 50.0 from https://matharena.ai/?comp=hmmt--hmmt_feb_2025 deleted.]
o3-mini-high,ifeval,93.9,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o3-mini-high=93.9.
o3-mini-high,livecodebench,68.8,https://arxiv.org/abs/2504.21318,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,Phi-4-reasoning paper Table 1 (o3-mini-high column): livecodebench=68.8. [R5d: prior unverified value 74.1 from https://openai.com/index/openai-o3-mini/ deleted.]
o3-mini-high,mmlu_pro,82.4,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: o3-mini-high=82.4.
o3-mini-high,swe_bench_verified,61.0,https://openai.com/index/introducing-gpt-4-5/,"{""effort"":""high"",""harness"":""official (GPT-4.5 blog)"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,OpenAI GPT-4.5 blog: o3-mini-high=61.0. (GPT-4.1 blog 477-subset value 49.3 dropped — was different N).
o3-mini-high,arc_agi_1,25.8,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: o3-mini-high=25.8.
o3-mini-high,arc_agi_2,3.0,https://arcprize.org/leaderboard,"{""effort"":""high"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""CoT"",""tools"":""none""}",True,leaderboard,verified,arcprize.org leaderboard audit: o3-mini (High) on leaderboard
o3-mini-high,usamo_2025,2.08,https://matharena.ai/?comp=usamo--usamo_2025,"""reasoning_effort:high""",True,leaderboard,verified,
gpt-4.5,simpleqa,62.5,https://www.helicone.ai/blog/gpt-4.5-benchmarks,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gpt-4.5,mmlu,90.8,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.5=90.8.
gpt-4.5,gpqa_diamond,69.5,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.5=69.5.
gpt-4.5,swe_bench_verified,38.0,https://openai.com/index/introducing-gpt-4-5/,"{""effort"":""n/a"",""harness"":""official (GPT-4.5 blog)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,OpenAI GPT-4.5 blog: gpt-4.5=38.0. (GPT-4.1 blog 477-subset value 38.0 dropped — was different N).
gpt-4.5,aime_2024,36.7,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.5=36.7.
gpt-4.5,arena_hard,51.4,https://github.com/lmarena/arena-hard-auto,"""Arena-Hard-v2.0 Creative Writing, Ensemble GPT-4.1+Gemini-2.5-Pro judge""",True,leaderboard,verified,
gpt-4.5,arc_agi_1,10.3,https://arcprize.org/leaderboard,"{""effort"":""n/a"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""non-thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""Base LLM"",""tools"":""none""}",True,leaderboard,verified,arcprize.org leaderboard audit: GPT-4.5 Base LLM on leaderboard
gpt-4.5,arc_agi_2,0.8,https://arcprize.org/leaderboard,"{""effort"":""n/a"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""non-thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""Base LLM"",""tools"":""none""}",True,leaderboard,verified,arcprize.org leaderboard audit: GPT-4.5 Base LLM on leaderboard
gpt-4.5,humaneval,86.6,https://www.helicone.ai/blog/gpt-4.5-benchmarks,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gpt-4.5,ifeval,88.2,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.5=88.2.
gpt-4.5,mmlu_pro,74.3,https://www.helicone.ai/blog/gpt-4.5-benchmarks,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gpt-4.1,swe_bench_verified,54.6,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,dropped,"OpenAI GPT-5 dev blog: gpt-4.1=54.6 (high reasoning effort). DROPPED: source uses 477/500 subset (omit 23 problems), not full SWE-bench Verified."
gpt-4.1,mmlu,90.2,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1=90.2.
gpt-4.1,gpqa_diamond,66.3,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1=66.3 (high reasoning effort).
gpt-4.1,ifeval,87.4,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1=87.4.
gpt-4.1,mmmu,74.8,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1=74.8 (high reasoning effort).
gpt-4.1,arena_hard,61.5,https://github.com/lmarena/arena-hard-auto,"""Arena-Hard-v2.0 Creative Writing, Ensemble GPT-4.1+Gemini-2.5-Pro judge""",True,leaderboard,verified,
gpt-4.1,aime_2024,48.1,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1=48.1.
gpt-4.1,aime_2025,46.4,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1=46.4 (high reasoning effort).
gpt-4.1,hle,5.4,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1=5.4 (high reasoning effort).
gpt-4.1,hmmt_feb_2025,28.9,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1=28.9 (high reasoning effort).
gpt-4.1,humaneval,92.0,https://openai.com/index/gpt-4-1/,{},,,dropped, [R5g: ghost cell — openai.com/gpt-4-1 blog does NOT contain HumanEval data; score 92.0 origin unknown. Dropped per R5g.]
gpt-4.1,livecodebench,44.7,https://arxiv.org/abs/2507.20534,"{""context"":""8192 max output tokens"",""effort"":""default"",""harness"":""Kimi team evaluation (official API)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unified (Kimi team setting)"",""tools"":""none (benchmark-specified)""}",True,tech_report,verified,"Via third-party self-test by Moonshot AI (Kimi K2 tech report, Table 3, p.16). LiveCodeBench v6 (Aug 2024–May 2025), non-thinking, 8192 max output tokens."
gpt-4.1,math_500,92.4,https://arxiv.org/abs/2507.20534,"{""context"":""8192 max output tokens"",""effort"":""default"",""harness"":""Kimi team evaluation (official API)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unified (Kimi team setting)"",""tools"":""none (benchmark-specified)""}",True,tech_report,verified,"Via third-party self-test by Moonshot AI (Kimi K2 tech report, Table 3, p.16). MATH-500 Acc, non-thinking, 8192 max output tokens."
gpt-4.1,mmlu_pro,81.8,https://arxiv.org/abs/2507.20534,"{""context"":""8192 max output tokens"",""effort"":""default"",""harness"":""Kimi team evaluation (official API)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (EM)"",""temperature"":""unified (Kimi team setting)"",""tools"":""none (benchmark-specified)""}",True,tech_report,verified,"Via third-party self-test by Moonshot AI (Kimi K2 tech report, Table 3, p.16). MMLU-Pro EM, non-thinking, 8192 max output tokens."
gpt-4.1,simpleqa,42.3,https://arxiv.org/abs/2507.20534,"{""context"":""8192 max output tokens"",""effort"":""default"",""harness"":""Kimi team evaluation (official API)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unified (Kimi team setting)"",""tools"":""none (benchmark-specified)""}",True,tech_report,verified,"Via third-party self-test by Moonshot AI (Kimi K2 tech report, Table 3, p.16). SimpleQA Correct, non-thinking, 8192 max output tokens."
gpt-4.1,arc_agi_1,5.5,https://arcprize.org/leaderboard,"{""effort"":""n/a"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""non-thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""Base LLM"",""tools"":""none""}",True,leaderboard,verified,arcprize.org leaderboard audit: GPT-4.1 Base LLM on leaderboard
gpt-4.1,arc_agi_2,0.4,https://arcprize.org/leaderboard,"{""effort"":""n/a"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""non-thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""Base LLM"",""tools"":""none""}",True,leaderboard,verified,arcprize.org leaderboard audit: GPT-4.1 Base LLM on leaderboard
gpt-4.1,codeforces_rating,1807.0,https://openai.com/index/gpt-4-1/,{},,,dropped, [R5g: ghost cell — openai.com/gpt-4-1 blog does NOT contain Codeforces data; score 1807 origin unknown. Dropped per R5g.]
gpt-4.1,mrcr_v2,80.0,https://awesomeagents.ai/leaderboards/long-context-benchmarks-leaderboard/,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gpt-4.1,terminal_bench_1,30.3,https://www.tbench.ai/leaderboard/terminal-bench/1.0,{},,leaderboard,verified,"Confirmed 30.3% on tbench.ai Terminal-Bench 1.0 leaderboard (Terminus 1 scaffold, 2025-05-15). model=gpt-4.1"
gpt-4.1-mini,ifeval,84.1,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1-mini=84.1.
gpt-4.1-mini,mmmu,72.7,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-mini=72.7 (high reasoning effort).
gpt-4.1-mini,arena_hard,28.2,https://github.com/lmarena/arena-hard-auto,"""Arena-Hard-v2.0 Creative Writing, Ensemble GPT-4.1+Gemini-2.5-Pro judge""",True,leaderboard,verified,
gpt-4.1-mini,gpqa_diamond,65.0,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-mini=65.0 (high reasoning effort).
gpt-4.1-mini,humaneval,90.0,https://www.helicone.ai/blog/gpt-4.1-full-developer-guide,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gpt-4.1-mini,mmlu,87.5,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1-mini=87.5.
gpt-4.1-mini,simpleqa,20.0,https://www.helicone.ai/blog/gpt-4.1-full-developer-guide,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gpt-4.1-mini,arc_agi_1,3.5,https://arcprize.org/leaderboard,"{""effort"":""n/a"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""non-thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""Base LLM"",""tools"":""none""}",True,leaderboard,verified,arcprize.org leaderboard audit: GPT-4.1-Mini Base LLM on leaderboard
gpt-4.1-mini,arc_agi_2,0.0,https://arcprize.org/leaderboard,"{""effort"":""n/a"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""non-thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""Base LLM"",""tools"":""none""}",True,leaderboard,verified,arcprize.org leaderboard audit: GPT-4.1-Mini Base LLM on leaderboard
gpt-4.1-mini,bigcodebench,48.9,https://bigcode-bench.github.io/,"""BigCodeBench Complete Instruct pass@1""",True,leaderboard,verified,
gpt-4.1-nano,mmlu,80.1,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1-nano=80.1.
gpt-4.1-nano,gpqa_diamond,50.3,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-nano=50.3 (high reasoning effort).
gpt-4.1-nano,arena_hard,10.7,https://github.com/lmarena/arena-hard-auto,"""Arena-Hard-v2.0 Creative Writing, Ensemble GPT-4.1+Gemini-2.5-Pro judge""",True,leaderboard,verified,
gpt-4.1-nano,humaneval,78.0,https://www.datacamp.com/blog/gpt-4-1,{},,,dropped," [Dropped: DataCamp is a tutorial/learning platform aggregator, not primary source. Rule R5h-tutorial-blog.]"
gpt-4.1-nano,ifeval,74.5,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1-nano=74.5.
gpt-4.1-nano,arc_agi_1,0.0,https://arcprize.org/leaderboard,"{""effort"":""n/a"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""non-thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""Base LLM"",""tools"":""none""}",True,leaderboard,verified,arcprize.org leaderboard audit: GPT-4.1-Nano Base LLM on leaderboard
gpt-4.1-nano,arc_agi_2,0.0,https://arcprize.org/leaderboard,"{""effort"":""n/a"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""non-thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""Base LLM"",""tools"":""none""}",True,leaderboard,verified,arcprize.org leaderboard audit: GPT-4.1-Nano Base LLM on leaderboard
o3-high,aime_2024,96.7,https://openai.com/index/introducing-o3-and-o4-mini/,{},,,needs_review,
o3-high,aime_2025,88.9,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o3-high=88.9 (high reasoning effort).
o3-high,gpqa_diamond,83.3,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o3-high=83.3 (high reasoning effort).
o3-high,swe_bench_verified,69.1,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,dropped,"OpenAI GPT-5 dev blog: o3-high=69.1 (high reasoning effort). DROPPED: source uses 477/500 subset (omit 23 problems), not full SWE-bench Verified."
o3-high,frontiermath,15.8,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""with python tool (per OpenAI dev blog)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""python tool""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o3-high=15.8 (high reasoning effort).
o3-high,livecodebench,75.8,https://www.alphaxiv.org/benchmarks/uc-berkeley/livecodebench,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
o3-high,math_500,98.1,https://github.com/openai/simple-evals,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"openai/simple-evals README (no tools, [^10]): o3-high=98.1 [fn6: MATH=MATH-500 for post-o1 models]. Prior ref (artificialanalysis.ai) reported 99.2 — overwritten."
o3-high,codeforces_rating,2706.0,https://www.datacamp.com/blog/o4-mini,{},,,dropped," [Dropped: DataCamp is a tutorial/learning platform aggregator, not primary source. Rule R5h-tutorial-blog.]"
o3-high,arc_agi_2,6.5,https://arcprize.org/leaderboard,"{""effort"":""high"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""CoT"",""tools"":""none""}",True,leaderboard,verified,arcprize.org leaderboard audit: o3 (High) on leaderboard
o3-high,arena_hard,85.9,https://llm-stats.com/benchmarks/arena-hard,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
o3-high,mmlu,93.3,https://github.com/openai/simple-evals,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"openai/simple-evals README (no tools, [^10]): o3-high=93.3."
o3-high,humaneval,88.4,https://github.com/openai/simple-evals,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"openai/simple-evals README (no tools, [^10]): o3-high=88.4."
o3-high,simpleqa,48.6,https://github.com/openai/simple-evals,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"openai/simple-evals README (no tools, [^10]): o3-high=48.6."
o3-high,hle,20.2,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o3-high=20.2 (high reasoning effort).
o3-high,browsecomp,49.7,https://www.helicone.ai/blog/o3-and-o4-mini-for-developers,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
o3-high,mmmu,82.9,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o3-high=82.9 (high reasoning effort).
o3-high,arc_agi_1,60.8,https://arcprize.org/leaderboard,"{""effort"":""high"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""CoT"",""tools"":""none""}",True,leaderboard,verified,arcprize.org leaderboard audit: o3 (High) on leaderboard
o3-high,hmmt_feb_2025,81.7,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o3-high=81.7 (high reasoning effort).
o3-high,ifeval,92.1,https://huggingface.co/Qwen/Qwen3-235B-A22B-Thinking-2507,{},,,dropped,
o3-high,imo_2025,16.67,https://matharena.ai/?comp=imo--imo_2025,"""reasoning_effort:high""",True,leaderboard,verified,
o3-high,mmlu_pro,85.9,https://huggingface.co/Qwen/Qwen3-235B-A22B-Thinking-2507,{},,,dropped,
o3-high,mmmu_pro,76.4,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""avg across standard and vision sets"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o3-high=76.4 (high reasoning effort).
o3-high,usamo_2025,21.9,https://files.sri.inf.ethz.ch/matharena/usamo_report.pdf,{},,,dropped,
o3-high,aa_lcr,69.0,https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning,"{""effort"":""high (AA default)"",""harness"":""AA standard evaluation"",""mode"":""thinking"",""notes"":""AA evaluates o3 at high effort by default per aa/evaluations/artificial-analysis-long-context-reasoning"",""sampling"":""pass@1""}",,third_party,verified,"ArtificialAnalysis aggregator; used as fallback when no primary source available (R5d-aggregator-fallback). AA LCR eval page: o3 (default=high effort) lcr=0.693 → 69.3%, displayed as 69. AA slug: o3."
o3-high,brumo_2025,95.83,https://matharena.ai/?comp=brumo--brumo_2025,"""reasoning_effort:high""",True,leaderboard,verified,
o3-high,cmimc_2025,79.38,https://matharena.ai/?comp=cmimc--cmimc_2025,"""reasoning_effort:high""",True,leaderboard,verified,
o3-high,critpt,1.4,https://github.com/CritPt-Benchmark/CritPt,"""CritPt leaderboard, \u03b5=1.5, % tasks solved""",True,leaderboard,verified,
o3-high,ifbench,69.3,https://github.com/allenai/IFBench,"""IFBench score, 0-shot""",True,leaderboard,verified,"Source lists ""OpenAI o3"" without explicit effort level"
o3-high,smt_2025,87.74,https://matharena.ai/?comp=smt--smt_2025,"""reasoning_effort:high""",True,leaderboard,verified,
o3-high,terminal_bench_1,30.2,https://www.tbench.ai/leaderboard/terminal-bench/1.0,{},,leaderboard,verified,"Confirmed 30.2% on tbench.ai Terminal-Bench 1.0 leaderboard (Terminus 1 scaffold, 2025-05-15). model listed as o3 (high effort)"
o4-mini-high,aime_2024,93.4,https://www.datacamp.com/blog/o4-mini,{},,,dropped," [Dropped: DataCamp is a tutorial/learning platform aggregator, not primary source. Rule R5h-tutorial-blog.]"
o4-mini-high,aime_2025,92.7,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o4-mini-high=92.7 (high reasoning effort).
o4-mini-high,gpqa_diamond,81.4,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o4-mini-high=81.4 (high reasoning effort).
o4-mini-high,swe_bench_verified,68.1,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,dropped,"OpenAI GPT-5 dev blog: o4-mini-high=68.1 (high reasoning effort). DROPPED: source uses 477/500 subset (omit 23 problems), not full SWE-bench Verified."
o4-mini-high,frontiermath,15.4,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""with python tool (per OpenAI dev blog)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""python tool""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o4-mini-high=15.4 (high reasoning effort).
o4-mini-high,livecodebench,80.2,https://www.alphaxiv.org/benchmarks/uc-berkeley/livecodebench,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
o4-mini-high,humaneval,99.3,https://github.com/openai/simple-evals,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"openai/simple-evals README (no tools, [^10]): o4-mini-high=99.3. Prior ref (rdworldonline) reported 98.5 — overwritten."
o4-mini-high,codeforces_rating,2719.0,https://www.datacamp.com/blog/o4-mini,{},,,dropped," [Dropped: DataCamp is a tutorial/learning platform aggregator, not primary source. Rule R5h-tutorial-blog.]"
o4-mini-high,arena_hard,79.1,https://llm-stats.com/benchmarks/arena-hard,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
o4-mini-high,mmlu,90.3,https://github.com/openai/simple-evals,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"openai/simple-evals README (no tools, [^10]): o4-mini-high=90.3."
o4-mini-high,math_500,98.2,https://github.com/openai/simple-evals,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"openai/simple-evals README (no tools, [^10]): o4-mini-high=98.2 [fn6: MATH=MATH-500]."
o4-mini-high,simpleqa,19.3,https://github.com/openai/simple-evals,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"openai/simple-evals README (no tools, [^10]): o4-mini-high=19.3."
o4-mini-high,hle,14.7,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o4-mini-high=14.7 (high reasoning effort).
o4-mini-high,arc_agi_2,6.1,https://arcprize.org/leaderboard,"{""effort"":""high"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""CoT"",""tools"":""none""}",True,leaderboard,verified,arcprize.org leaderboard audit: o4-mini (High) on leaderboard
o4-mini-high,mmmu,81.6,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o4-mini-high=81.6 (high reasoning effort).
o4-mini-high,hmmt_feb_2025,85.0,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o4-mini-high=85.0 (high reasoning effort).
o4-mini-high,ifeval,92.4,https://huggingface.co/Qwen/Qwen3-235B-A22B-Thinking-2507,{},True,model_card,verified,
o4-mini-high,mmlu_pro,81.9,https://huggingface.co/Qwen/Qwen3-235B-A22B-Thinking-2507,{},,,dropped,
o4-mini-high,usamo_2025,19.3,https://files.sri.inf.ethz.ch/matharena/usamo_report.pdf,{},,,dropped,
o4-mini-high,arc_agi_1,58.7,https://arcprize.org/leaderboard,"{""effort"":""high"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""CoT"",""tools"":""none""}",True,leaderboard,verified,arcprize.org leaderboard audit: o4-mini (High) on leaderboard
o4-mini-high,brumo_2025,86.67,https://matharena.ai/?comp=brumo--brumo_2025,"""reasoning_effort:high""",True,leaderboard,verified,
o4-mini-high,cmimc_2025,84.38,https://matharena.ai/?comp=cmimc--cmimc_2025,"""reasoning_effort:high""",True,leaderboard,verified,
o4-mini-high,critpt,0.6,https://github.com/CritPt-Benchmark/CritPt,"""CritPt leaderboard, \u03b5=1.5, % tasks solved""",True,leaderboard,verified,
o4-mini-high,imo_2025,14.29,https://matharena.ai/?comp=imo--imo_2025,"""reasoning_effort:high""",True,leaderboard,verified,
o4-mini-high,smt_2025,88.68,https://matharena.ai/?comp=smt--smt_2025,"""reasoning_effort:high""",True,leaderboard,verified,
o4-mini-high,terminal_bench_1,18.5,https://www.tbench.ai/leaderboard/terminal-bench/1.0,{},,leaderboard,verified,"Confirmed 18.5% on tbench.ai Terminal-Bench 1.0 leaderboard (Terminus 1 scaffold, 2025-05-15). model listed as o4-mini (high effort)"
gpt-5,aime_2025,94.6,https://openai.com/index/gpt-5-1-for-developers/,"{""effort"":""high"",""harness"":""official with apply_patch"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.1 dev blog (cross-model column): GPT-5 = 94.6.
gpt-5,gpqa_diamond,85.7,https://openai.com/index/gpt-5-1-for-developers/,"{""effort"":""high"",""harness"":""official with apply_patch"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.1 dev blog (cross-model column): GPT-5 = 85.7.
gpt-5,swe_bench_verified,74.9,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: gpt-5=74.9.
gpt-5,mmmu,84.2,https://openai.com/index/gpt-5-1-for-developers/,"{""effort"":""high"",""harness"":""official with apply_patch"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.1 dev blog (cross-model column): GPT-5 = 84.2.
gpt-5,math_500,99.4,https://artificialanalysis.ai/evaluations/math-500,"{""effort"":""high (AA default)"",""harness"":""AA standard evaluation"",""mode"":""thinking"",""notes"":""Per AA evaluations/math-500: gpt-5 math_500=99.4%"",""sampling"":""pass@1""}",,third_party,verified,"ArtificialAnalysis aggregator; used as fallback when no primary source available (R5d-aggregator-fallback). AA MATH-500 eval page: gpt-5 math_500=0.994 → 99.4%. AA slug: gpt-5 (deprecated, reasoning variant)."
gpt-5,tau_bench_telecom,97.0,https://www.digitalocean.com/resources/articles/gpt-5-overview,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gpt-5,hle,24.8,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-5=24.8 (high reasoning effort).
gpt-5,browsecomp,54.9,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: gpt-5=54.9.
gpt-5,mmlu_pro,87.5,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: gpt-5=87.5.
gpt-5,aa_lcr,76.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2 model card: gpt-5=76.
gpt-5,simpleqa,55.0,https://cdn.openai.com/gpt-5-system-card.pdf,{},True,official_paper,verified,
gpt-5,frontiermath,26.3,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""with python tool (per OpenAI dev blog)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""python tool""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-5=26.3 (high reasoning effort).
gpt-5,livecodebench,84.5,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: gpt-5=84.5.
gpt-5,codeforces_rating,2537.0,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: gpt-5=2537.
gpt-5,swe_bench_pro,41.78,https://scale.com/leaderboard/swe_bench_pro_public,{},False,leaderboard,verified, [Verified against Scale.com SWE-Bench Pro Public leaderboard (2026-04-29): gpt-5-2025-08-07 (High) = 41.78. Scale uses standardized evaluation scaffolding.]
gpt-5,arc_agi_1,65.7,https://arcprize.org/leaderboard,"{""effort"":""high"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""CoT"",""tools"":""none""}",True,leaderboard,verified,arcprize.org leaderboard audit: GPT-5 (High) on leaderboard
gpt-5,arc_agi_2,9.9,https://arcprize.org/leaderboard,"{""effort"":""high"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""CoT"",""tools"":""none""}",True,leaderboard,verified,arcprize.org leaderboard audit: GPT-5 (High) on leaderboard
gpt-5,chatbot_arena_elo,1460.0,https://lmarena.ai/,{},,,needs_review,
gpt-5,humaneval,94.0,https://cdn.openai.com/gpt-5-system-card.pdf,{},,,dropped,
gpt-5,ifeval,90.0,https://cdn.openai.com/gpt-5-system-card.pdf,{},,,dropped,
gpt-5,mmlu,91.0,https://cdn.openai.com/gpt-5-system-card.pdf,{},,,dropped,
gpt-5,osworld,40.0,https://www.digitalocean.com/resources/articles/gpt-5-overview,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gpt-5,terminal_bench,35.2,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: gpt-5=35.2.
gpt-5,aa_intelligence_index,69.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2 model card: gpt-5=69.
gpt-5,aime_2024,94.6,https://openai.com/index/introducing-gpt-5/,{},,,dropped,
gpt-5,brumo_2025,91.67,https://matharena.ai/?comp=brumo--brumo_2025,"""reasoning_effort:high""",True,leaderboard,verified,
gpt-5,cmimc_2025,90.0,https://matharena.ai/?comp=cmimc--cmimc_2025,"""reasoning_effort:high""",True,leaderboard,verified,
gpt-5,critpt,5.7,https://github.com/CritPt-Benchmark/CritPt,"""CritPt leaderboard, \u03b5=1.5, % tasks solved""",True,leaderboard,verified,
gpt-5,hmmt_nov_2025,89.2,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: gpt-5=89.2.
gpt-5,imo_2025,38.1,https://matharena.ai/?comp=imo--imo_2025,"""reasoning_effort:high""",True,leaderboard,verified,
gpt-5,livebench,71.3,https://livebench.ai/,"{""effort"":""n/a"",""mode"":""thinking"",""tools"":""n/a""}",False,leaderboard,verified,"livebench.ai table_2026_01_08 (latest as of 2026-04-28): gpt-5-pro-2025-10-06 avg=71.3. Matched as closest GPT-5 entry; canonical effort=high not confirmed for this variant (no ""-high"" suffix). See review. BP had 70.5."
gpt-5,matharena_apex_2025,1.04,https://matharena.ai/?comp=apex--apex_2025,"""reasoning_effort:high""",True,leaderboard,verified,
gpt-5,simplebench,61.6,https://lmcouncil.ai/benchmarks,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail | DROPPED (R5g-c stale, never verified, source not audited in any pass)"
gpt-5,smt_2025,91.98,https://matharena.ai/?comp=smt--smt_2025,"""reasoning_effort:high""",True,leaderboard,verified,
gpt-5,terminal_bench_1,41.3,https://www.tbench.ai/leaderboard/terminal-bench/1.0,{},,leaderboard,verified,"Confirmed 41.3% on tbench.ai Terminal-Bench 1.0 leaderboard (Terminus 2 scaffold, 2025-08-11)."
gpt-oss-120b,aime_2025,97.9,https://arxiv.org/abs/2508.10925,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""harmony"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""with tools""}",True,tech_report,verified,gpt-oss model card Table 3 (high effort): AIME 2025 with-tools 97.9
gpt-oss-120b,gpqa_diamond,80.9,https://arxiv.org/abs/2508.10925,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""harmony"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""with tools""}",True,tech_report,verified,gpt-oss model card Table 3 (high effort): GPQA Diamond with-tools 80.9
gpt-oss-120b,mmlu_pro,90.0,https://www.clarifai.com/blog/openai-gpt-oss-benchmarks,{},,,dropped, | DROPPED: BP value 90.0 from clarifai blog matches gpt-oss-120b MMLU=90.0 in Table 3 — clarifai mislabeled MMLU as MMLU-Pro. Table 3 does not report MMLU-Pro for gpt-oss. Ghost cell per R5g(b).
gpt-oss-120b,codeforces_rating,2622.0,https://arxiv.org/abs/2508.10925,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""harmony"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""with tools""}",True,tech_report,verified,gpt-oss model card Table 3 (high effort): Codeforces with-tools high 2622 (was 2620). [R5d: prior unverified value 2620 from https://smythos.com/developers/ai-models/openai-gpt-oss-120b-and-20b/ deleted.]
gpt-oss-120b,tau_bench_retail,67.8,https://arxiv.org/abs/2508.10925,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""harmony"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""with tools""}",True,tech_report,verified,gpt-oss model card Table 3 (high effort): Tau-Bench Retail with-tools 67.8
gpt-oss-120b,mmlu,90.0,https://arxiv.org/abs/2508.10925,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""harmony"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""with tools""}",True,tech_report,verified,gpt-oss model card Table 3 (high effort): MMLU 90.0
gpt-oss-120b,swe_bench_verified,62.4,https://arxiv.org/abs/2508.10925,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""harmony"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""with tools""}",True,tech_report,verified,gpt-oss model card Table 3 (high effort): SWE-Bench Verified 62.4
gpt-oss-120b,aime_2024,96.6,https://arxiv.org/abs/2508.10925,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""harmony"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""with tools""}",True,tech_report,verified,gpt-oss model card Table 3 (high effort): AIME 2024 with-tools 96.6
gpt-oss-120b,humaneval,92.0,https://arxiv.org/abs/2508.10925,{},,,dropped,
gpt-oss-120b,ifeval,88.0,https://arxiv.org/abs/2508.10925,{},,,dropped,
gpt-oss-120b,livecodebench,75.0,https://arxiv.org/abs/2508.10925,{},,,dropped,
gpt-oss-120b,math_500,98.0,https://www.clarifai.com/blog/openai-gpt-oss-benchmarks,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gpt-oss-120b,terminal_bench,18.7,https://www.tbench.ai/leaderboard/terminal-bench/2.0,{},,leaderboard,verified,"Confirmed 18.7% on tbench.ai Terminal-Bench 2.0 leaderboard (Terminus 2 scaffold, 2025-11-01)."
gpt-oss-20b,aime_2025,98.7,https://arxiv.org/abs/2508.10925,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""harmony"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""with tools""}",True,tech_report,verified,gpt-oss model card Table 3 (high effort): AIME 2025 with-tools 98.7
gpt-oss-20b,gpqa_diamond,71.5,https://arxiv.org/abs/2508.10925,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""harmony"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,gpt-oss model card Table 3 (high effort): GPQA Diamond no-tools 71.5
gpt-oss-20b,mmlu_pro,85.3,https://www.clarifai.com/blog/openai-gpt-oss-benchmarks,{},,,dropped, | DROPPED: BP value 85.3 from clarifai blog matches gpt-oss-20b MMLU=85.3 in Table 3 — clarifai mislabeled MMLU as MMLU-Pro. Ghost cell per R5g(b).
gpt-oss-20b,mmlu,85.3,https://arxiv.org/abs/2508.10925,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""harmony"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""with tools""}",True,tech_report,verified,gpt-oss model card Table 3 (high effort): MMLU 85.3
gpt-oss-20b,ifeval,86.73,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 86.73. Official Liquid AI reported result.
gpt-oss-20b,livecodebench,70.0,https://arxiv.org/abs/2508.10925,{},,,dropped,
gpt-oss-20b,math_500,92.4,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 92.40. Official Liquid AI reported result.
gpt-oss-20b,terminal_bench,3.1,https://www.tbench.ai/leaderboard/terminal-bench/2.0,{},,leaderboard,verified,"Confirmed 3.1% on tbench.ai Terminal-Bench 2.0 leaderboard (Terminus 2 scaffold, 2025-11-01)."
gpt-oss-20b,aime_2024,96.0,https://arxiv.org/abs/2508.10925,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""harmony"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""with tools""}",True,tech_report,verified,"gpt-oss model card Table 3 (high effort): AIME 2024 with-tools 96.0 (was 98.7 from llm-stats — wrong, that was AIME 2025 score). [R5d: prior unverified value 98.7 from https://llm-stats.com/models/compare/gpt-oss-120b-vs-gpt-oss-20b deleted.]"
gpt-oss-20b,codeforces_rating,2516.0,https://arxiv.org/abs/2508.10925,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""harmony"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""with tools""}",True,tech_report,verified,gpt-oss model card Table 3 (high effort): Codeforces with-tools high 2516 (was 1985 — likely no-tools or low effort). [R5d: prior unverified value 1985 from https://arxiv.org/html/2508.10925v1 deleted.]
gpt-oss-20b,humaneval,85.0,https://arxiv.org/html/2508.10925v1,{},,,dropped,
gpt-oss-20b,swe_bench_verified,60.7,https://arxiv.org/abs/2508.10925,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""harmony"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""with tools""}",True,tech_report,verified,gpt-oss model card Table 3 (high effort): SWE-Bench Verified with-tools high 60.7 (was 52 from clarifai — likely medium effort). [R5d: prior unverified value 52 from https://www.clarifai.com/blog/openai-gpt-oss-benchmarks-how-it-compares-to-glm-4.5-qwen3-deepseek-and-kimi-k2 deleted.]
gpt-5.1,aime_2025,94.0,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.2 blog (cross-model column): GPT-5.1 = 94.0.
gpt-5.1,gpqa_diamond,88.1,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.2 blog (cross-model column): GPT-5.1 = 88.1.
gpt-5.1,frontiermath,31.0,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.2 blog (cross-model column): GPT-5.1 = 31.0.
gpt-5.1,arc_agi_2,17.6,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.2 blog (cross-model column): GPT-5.1 = 17.6.
gpt-5.1,swe_bench_verified,76.3,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,OpenAI GPT-5.2 blog (cross-model column): GPT-5.1 = 76.3.
gpt-5.1,aa_lcr,75.0,https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning,"{""effort"":""default"",""harness"":""AA standard evaluation"",""mode"":""thinking"",""notes"":""Per AA evaluations/artificial-analysis-long-context-reasoning: gpt-5-1 lcr=75.0%"",""sampling"":""pass@1""}",,third_party,verified,ArtificialAnalysis aggregator; used as fallback when no primary source available (R5d-aggregator-fallback). AA LCR eval page: gpt-5-1 lcr=0.750 → 75.0%. AA slug: gpt-5-1.
gpt-5.1,chatbot_arena_elo,1439.0,https://lmarena.ai/leaderboard/text,{},True,leaderboard,verified,lmarena.ai text arena (fetched 2026-04-29). Arena model name: gpt-5.1 (rank 52). ELO updates continuously; score reflects latest available.
gpt-5.1,mmmu_pro,85.4,https://www.vellum.ai/blog/gpt-5-2-benchmarks,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gpt-5.1,browsecomp,50.8,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.2 blog (cross-model column): GPT-5.1 = 50.8.
gpt-5.1,hle,25.7,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (no-tools column)""}",True,official_blog,verified,OpenAI GPT-5.2 blog (cross-model column): GPT-5.1 = 25.7.
gpt-5.1,livecodebench,82.0,https://www.vellum.ai/blog/flagship-model-report,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gpt-5.1,math_500,99.0,https://artificialanalysis.ai/evaluations/math-500,{},,,needs_review,"Reference URL is AA math-500 eval page, but gpt-5-1 (gpt-5.1) is NOT present on that page; AA shows math_500=null for this model. Possible ghost score (R5g-a). Cannot verify 99.0% from AA source."
gpt-5.1,matharena_apex_2025,1.04,https://matharena.ai/?comp=apex--apex_2025,"""reasoning_effort:high""",True,leaderboard,verified,
gpt-5.1,mmlu,90.0,https://www.vellum.ai/blog/gpt-5-2-benchmarks,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gpt-5.1,mmmu,85.4,https://openai.com/index/gpt-5-1-for-developers/,"{""effort"":""high"",""harness"":""official with apply_patch"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.1 dev blog: 85.4 (high reasoning effort).
gpt-5.1,simpleqa,55.0,https://www.vellum.ai/blog/gpt-5-2-benchmarks,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gpt-5.1,swe_bench_pro,50.8,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,OpenAI GPT-5.2 blog (cross-model column): GPT-5.1 = 50.8.
gpt-5.1,terminal_bench,47.6,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-4.7 blog: gpt-5.1=47.6.
gpt-5.1,arc_agi_1,72.8,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.2 blog (cross-model column): GPT-5.1 = 72.8.
gpt-5.1,brumo_2025,93.33,https://matharena.ai/?comp=brumo--brumo_2025,"""reasoning_effort:high""",True,leaderboard,verified,
gpt-5.1,cmimc_2025,91.88,https://matharena.ai/?comp=cmimc--cmimc_2025,"""reasoning_effort:high""",True,leaderboard,verified,
gpt-5.1,hmmt_nov_2025,91.67,https://matharena.ai/?comp=hmmt--hmmt_nov_2025,"""reasoning_effort:high""",True,leaderboard,verified,
gpt-5.1,mmlu_pro,87.0,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-4.7 blog: gpt-5.1=87.0.
gpt-5.1,smt_2025,91.04,https://matharena.ai/?comp=smt--smt_2025,"""reasoning_effort:high""",True,leaderboard,verified,
gpt-5.2,aime_2025,100.0,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.2 blog: 100.0 (xhigh reasoning effort).
gpt-5.2,gpqa_diamond,92.4,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.2 blog: 92.4 (xhigh reasoning effort).
gpt-5.2,frontiermath,40.3,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.2 blog: 40.3 (xhigh reasoning effort).
gpt-5.2,swe_bench_verified,80.0,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,OpenAI GPT-5.2 blog: 80.0 (xhigh reasoning effort).
gpt-5.2,swe_bench_pro,55.6,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,OpenAI GPT-5.2 blog: 55.6 (xhigh reasoning effort).
gpt-5.2,arc_agi_2,52.9,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.2 blog: 52.9 (xhigh reasoning effort).
gpt-5.2,browsecomp,65.8,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.2 blog: 65.8 (xhigh reasoning effort).
gpt-5.2,critpt,11.6,https://artificialanalysis.ai/evaluations/critpt,"{""effort"":""xhigh (AA default for gpt-5-2)"",""harness"":""AA standard evaluation"",""mode"":""thinking"",""notes"":""Per AA evaluations/critpt: gpt-5-2 critpt=11.6%"",""sampling"":""pass@1""}",,third_party,verified,ArtificialAnalysis aggregator; used as fallback when no primary source available (R5d-aggregator-fallback). AA CritPT eval page: gpt-5-2 critpt=0.116 → 11.6%. AA slug: gpt-5-2.
gpt-5.2,scicode,52.0,https://deepmind.google/models/gemini/pro/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/pro/ Performance table (PROMOTED from candidate; prior unverified value 54.6 from https://artificialanalysis.ai/evaluations/scicode deleted per audit rule). DeepMind /models/gemini/pro/ cross-model Performance table: SciCode 52%
gpt-5.2,aa_lcr,72.3,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: gpt-5.2=72.3.
gpt-5.2,tau_bench_telecom,98.7,https://openai.com/index/introducing-gpt-5-2/,{},,,dropped,
gpt-5.2,mmmu_pro,79.5,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.2 blog: 79.5 (xhigh reasoning effort).
gpt-5.2,simpleqa,58.0,https://llm-stats.com/models/gpt-5.2-2025-12-11,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gpt-5.2,ifeval,94.8,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,model_card,verified,"Qwen3.5-397B-A17B HF model card cross-model table (Qwen self-test of 3rd-party model, mc=false): ifeval 94.8 (3rd-party Qwen self-test). [R5d: prior unverified value 95.0 from https://llm-stats.com/models/gpt-5.2-2025-12-11 deleted.]"
gpt-5.2,livecodebench,87.7,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,model_card,verified,"Qwen3.5-397B-A17B HF model card cross-model table (Qwen self-test of 3rd-party model, mc=false): livecodebench 87.7 (3rd-party Qwen self-test). [R5d: prior unverified value 80.0 from https://llm-stats.com/models/gpt-5.2-2025-12-11 deleted.]"
gpt-5.2,humaneval,95.0,https://llm-stats.com/models/gpt-5.2-2025-12-11,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gpt-5.2,mmlu,88.0,https://llm-stats.com/models/gpt-5.2-2025-12-11,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gpt-5.2,math_500,99.4,https://artificialanalysis.ai/evaluations/math-500,{},,,needs_review,"Reference URL is AA math-500 eval page, but gpt-5-2 (gpt-5.2) is NOT present on that page; AA shows math_500=null for this model. Possible ghost score (R5g-a). Cannot verify 99.4% from AA source."
gpt-5.2,mrcr_v2,83.8,https://deepmind.google/models/gemini/pro/,"{""context"":""128k"",""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/pro/ Performance table (PROMOTED from candidate; prior unverified value 70.0 from https://www.datacamp.com/blog/gpt-5-2 deleted per audit rule). DeepMind /models/gemini/pro/ cross-model Performance table: MRCR v2 8-needle 128k average 83.8%
gpt-5.2,osworld,47.3,https://openai.com/index/introducing-gpt-5-4/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.4 blog table: 47.3.
gpt-5.2,hle,34.5,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (no-tools column)""}",True,official_blog,verified,OpenAI GPT-5.2 blog: 34.5 (xhigh reasoning effort).
gpt-5.2,mmmu,86.7,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,model_card,verified,"Qwen3.5-397B-A17B HF model card cross-model table (Qwen self-test of 3rd-party model, mc=false): mmmu 86.7 (matches prior unverified value, re-sourced to HF Qwen3.5 card)."
gpt-5.2,chatbot_arena_elo,1438.0,https://lmarena.ai/leaderboard/text,{},True,leaderboard,verified,lmarena.ai text arena (fetched 2026-04-29). Arena model name: gpt-5.2 (rank 54). ELO updates continuously; score reflects latest available.
gpt-5.2,aa_intelligence_index,70.0,https://x.com/ArtificialAnlys/status/1943166841150644622,{},,,needs_review,
gpt-5.2,arc_agi_1,86.2,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.2 blog: 86.2 (xhigh reasoning effort).
gpt-5.2,codeforces_rating,3148.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gpt-5.2=3148.
gpt-5.2,mmlu_pro,86.7,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: gpt-5.2=86.7.
gpt-5.2,tau_bench_retail,88.0,https://openai.com/index/introducing-gpt-5-2/,{},,,dropped,
gpt-5.2,terminal_bench,62.2,https://openai.com/index/introducing-gpt-5-4/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.4 blog table: 62.2.
gpt-5.2,aime_2024,100.0,https://www.vellum.ai/blog/gpt-5-2-benchmarks,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gpt-5.2,brumo_2025,98.33,https://matharena.ai/?comp=brumo--brumo_2025,"""reasoning_effort:high""",False,leaderboard,verified,
gpt-5.2,cmimc_2025,91.25,https://matharena.ai/?comp=cmimc--cmimc_2025,"""reasoning_effort:high""",False,leaderboard,verified,
gpt-5.2,hmmt_nov_2025,97.1,https://z.ai/blog/glm-5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5 blog: gpt-5.2=97.1.
gpt-5.2,matharena_apex_2025,13.54,https://matharena.ai/?comp=apex--apex_2025,"""reasoning_effort:high""",False,leaderboard,verified,
gpt-5.2,smt_2025,91.98,https://matharena.ai/?comp=smt--smt_2025,"""reasoning_effort:high""",False,leaderboard,verified,
gpt-5.3-codex,swe_bench_pro,56.8,https://openai.com/index/introducing-gpt-5-3-codex/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,OpenAI GPT-5.3-Codex blog: 56.8 (xhigh reasoning effort).
gpt-5.3-codex,osworld,64.7,https://openai.com/index/introducing-gpt-5-3-codex/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.3-Codex blog: 64.7 (xhigh reasoning effort).
gpt-5.3-codex,swe_bench_verified,56.8,https://openai.com/index/introducing-gpt-5-3-codex/,{},,,dropped,
gpt-5.3-codex,humaneval,93.0,https://openai.com/index/introducing-gpt-5-3-codex/,{},,,dropped,
gpt-5.3-codex,aime_2025,94.0,https://automatio.ai/models/gpt-5-3-codex,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gpt-5.3-codex,gpqa_diamond,92.6,https://openai.com/index/introducing-gpt-5-4/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.4 blog table: 92.6.
gpt-5.3-codex,gsm8k,99.0,https://automatio.ai/models/gpt-5-3-codex,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gpt-5.3-codex,livecodebench,85.0,https://openai.com/index/introducing-gpt-5-3-codex/,{},,,dropped,
gpt-5.3-codex,math_500,96.0,https://automatio.ai/models/gpt-5-3-codex,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gpt-5.3-codex,mmlu,94.0,https://automatio.ai/models/gpt-5-3-codex,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gpt-5.3-codex,mmmu,84.0,https://automatio.ai/models/gpt-5-3-codex,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gpt-5.3-codex,terminal_bench,77.3,https://openai.com/index/introducing-gpt-5-3-codex/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.3-Codex blog: 77.3 (xhigh reasoning effort).
gpt-5.3-codex,ifeval,92.0,https://automatio.ai/models/gpt-5-3-codex,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gpt-5.4,swe_bench_pro,57.7,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 57.7.
gpt-5.4,terminal_bench,75.1,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 75.1.
gpt-5.4,mmmu_pro,81.2,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 81.2.
gpt-5.4,browsecomp,82.7,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 82.7.
gpt-5.4,frontiermath,47.6,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 47.6.
gpt-5.4,gpqa_diamond,92.8,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 92.8.
gpt-5.4,hle,39.8,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (no-tools column)""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 39.8.
gpt-5.4,arc_agi_1,93.7,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 93.7.
gpt-5.4,arc_agi_2,73.3,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 73.3.
gpt-5.5,swe_bench_pro,58.6,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 58.6.
gpt-5.5,terminal_bench,82.7,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 82.7.
gpt-5.5,mmmu_pro,81.2,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 81.2.
gpt-5.5,browsecomp,84.4,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 84.4.
gpt-5.5,frontiermath,51.7,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 51.7.
gpt-5.5,gpqa_diamond,93.6,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 93.6.
gpt-5.5,hle,41.4,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (no-tools column)""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 41.4.
gpt-5.5,arc_agi_1,95.0,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 95.0.
gpt-5.5,arc_agi_2,85.0,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 85.0.
claude-3.7-sonnet,gpqa_diamond,78.2,https://www.anthropic.com/news/claude-4,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (single-pass)"",""temperature"":""top_p=0.95"",""tools"":""none""}",True,official_blog,verified,Anthropic Claude 4 blog cross-model table: Sonnet 3.7 = 78.2% (with extended thinking).
claude-3.7-sonnet,aime_2024,23.3,https://www.anthropic.com/news/claude-3-7-sonnet,"{""effort"":null,""harness"":""official"",""judge"":""official harness"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":0.0,""tools"":""none""}",True,official_blog,verified,"Anthropic blog, no extended thinking. 80.0% (extended w/ parallel test-time compute) excluded."
claude-3.7-sonnet,math_500,82.2,https://www.anthropic.com/news/claude-3-7-sonnet,"{""effort"":null,""harness"":""official"",""judge"":""official harness"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":0.0,""tools"":""none""}",True,official_blog,verified,"Anthropic blog, no extended thinking. 96.2% (extended) excluded."
claude-3.7-sonnet,swe_bench_verified,62.3,https://www.anthropic.com/news/claude-4,"{""effort"":""default"",""harness"":""simple bash+editor scaffold"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 10 trials)"",""temperature"":""top_p=0.95"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,Anthropic Claude 4 blog cross-model table: Sonnet 3.7 = 62.3% / 70.3% w/ parallel.
claude-3.7-sonnet,ifeval,90.8,https://www.anthropic.com/news/claude-3-7-sonnet,"{""effort"":null,""harness"":""official"",""judge"":""official harness"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":0.0,""tools"":""none""}",True,official_blog,verified,"Anthropic blog, no extended thinking. 93.2% (extended) excluded."
claude-3.7-sonnet,humaneval,94.0,https://www.datacamp.com/blog/claude-3-7-sonnet,{},,,dropped," [Dropped: DataCamp is a tutorial/learning platform aggregator, not primary source. Rule R5h-tutorial-blog.]"
claude-3.7-sonnet,livecodebench,52.6,https://arxiv.org/abs/2504.21318,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,Phi-4-reasoning paper Table 1 (claude-3.7-sonnet column): livecodebench=52.6. [R5d: prior unverified value 65.0 from https://www.datacamp.com/blog/claude-3-7-sonnet deleted.]
claude-3.7-sonnet,tau_bench_retail,81.2,https://www.anthropic.com/news/claude-4,"{""effort"":""default"",""harness"":""\u03c4-bench official + max steps 100"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""prompt addendum to Agent Policy (Anthropic)"",""sampling"":""pass@1 (single-pass)"",""temperature"":""top_p=0.95"",""tools"":""agentic (TAU-bench official tools)""}",False,official_blog,verified,Anthropic Claude 4 blog cross-model table.
claude-3.7-sonnet,tau_bench_airline,58.4,https://www.anthropic.com/news/claude-4,"{""effort"":""default"",""harness"":""\u03c4-bench official + max steps 100"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""prompt addendum to Agent Policy (Anthropic)"",""sampling"":""pass@1 (single-pass)"",""temperature"":""top_p=0.95"",""tools"":""agentic (TAU-bench official tools)""}",False,official_blog,verified,Anthropic Claude 4 blog cross-model table.
claude-3.7-sonnet,tau_bench_telecom,49.0,https://artificialanalysis.ai/models/claude-3-7-sonnet,{},,,needs_review,Reference URL is AA model page (claude-3-7-sonnet). AA only publishes aggregate tau2 score (no domain breakdown); AA shows claude-3-7-sonnet tau2=0.5 aggregate but no telecom-domain score. Cannot verify 49.0 from AA source.
claude-3.7-sonnet,aime_2025,54.8,https://www.anthropic.com/news/claude-4,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (single-pass)"",""temperature"":""top_p=0.95"",""tools"":""none""}",True,official_blog,verified,Anthropic Claude 4 blog cross-model table.
claude-3.7-sonnet,frontiermath,3.1,https://epoch.ai/benchmarks/frontiermath,"{""effort"":""n/a"",""mode"":""non-thinking"",""tools"":""none""}",True,leaderboard,verified,"epoch.ai FrontierMath-2025-02-28-Private: claude-3-7-sonnet-20250219 (no thinking budget) = 3.1%. Canonical notes state no extended thinking for FrontierMath. BP had 5.0 (unverified, sourced from unknown)."
claude-3.7-sonnet,arc_agi_2,0.7,https://arcprize.org/leaderboard,"{""effort"":""16K"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""CoT"",""tools"":""none""}",True,leaderboard,verified,arcprize.org leaderboard audit: Claude 3.7 (16K) on leaderboard
claude-3.7-sonnet,mmmu,75.0,https://www.anthropic.com/news/claude-4,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (single-pass)"",""temperature"":""top_p=0.95"",""tools"":""none""}",True,official_blog,verified,Anthropic Claude 4 blog cross-model table (MMMU validation).
claude-3.7-sonnet,osworld,28.0,https://llm-stats.com/benchmarks/osworld,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
claude-3.7-sonnet,chatbot_arena_elo,1372.0,https://lmarena.ai/leaderboard/text,{},True,leaderboard,verified,lmarena.ai text arena (fetched 2026-04-29). Arena model name: claude-3-7-sonnet-20250219 (rank 138). ELO updates continuously; score reflects latest available.
claude-3.7-sonnet,arena_hard,63.9,https://github.com/lmarena/arena-hard-auto,"""Arena-Hard-v2.0 Creative Writing, Ensemble GPT-4.1+Gemini-2.5-Pro judge""",True,leaderboard,verified,
claude-3.7-sonnet,mmlu,89.0,https://www.anthropic.com/news/claude-3-7-sonnet,{},,,dropped,"DROPPED: Blog shows MMMLU (not MMLU), no MMLU row in source; 89.0 not found anywhere in blog table"
claude-3.7-sonnet,simpleqa,26.0,https://artificialanalysis.ai/models/claude-3-7-sonnet,{},,,needs_review,Reference URL is AA model page (claude-3-7-sonnet). AA does not publish SimpleQA scores; field is not present in AA model data for any model. Cannot verify from AA source. Original source needs identification.
claude-3.7-sonnet,arc_agi_1,28.6,https://arcprize.org/leaderboard,"{""effort"":""16K"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""CoT"",""tools"":""none""}",True,leaderboard,verified,arcprize.org leaderboard audit: Claude 3.7 (16K) on leaderboard
claude-3.7-sonnet,brumo_2025,65.83,https://matharena.ai/?comp=brumo--brumo_2025,"""extended_thinking""",True,leaderboard,verified,
claude-3.7-sonnet,codeforces_rating,1640.0,https://automatio.ai/models/claude-3-7-sonnet,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
claude-3.7-sonnet,mmlu_pro,78.0,https://automatio.ai/models/claude-3-7-sonnet,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
claude-3.7-sonnet,mmmlu,85.9,https://www.anthropic.com/news/claude-4,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""avg 14 non-English languages"",""sampling"":""pass@1 (single-pass)"",""temperature"":""top_p=0.95"",""tools"":""none""}",True,official_blog,verified,Anthropic Claude 4 blog cross-model table.
claude-3.7-sonnet,smt_2025,56.6,https://matharena.ai/?comp=smt--smt_2025,"""extended_thinking""",True,leaderboard,verified,
claude-3.7-sonnet,swe_bench_pro,25.6,https://www.anthropic.com/news/claude-3-7-sonnet,{},,,dropped,DROPPED: SWE-bench Pro not in blog table; only SWE-bench Verified is shown. Value 25.6 not in source.
claude-3.7-sonnet,terminal_bench_1,35.2,https://www.tbench.ai/leaderboard/terminal-bench/1.0,{},,leaderboard,verified,"Confirmed 35.2% on tbench.ai Terminal-Bench 1.0 leaderboard (Claude Code (Anthropic) scaffold, 2025-05-16)."
claude-3.7-sonnet,usamo_2025,3.65,https://matharena.ai/?comp=usamo--usamo_2025,"""extended_thinking""",True,leaderboard,verified,
claude-sonnet-4,gpqa_diamond,75.4,https://www.anthropic.com/news/claude-4,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (single-pass)"",""temperature"":""top_p=0.95"",""tools"":""none""}",True,official_blog,verified,Anthropic Claude 4 blog table: 75.4% single-pass / 83.8% w/ parallel.
claude-sonnet-4,swe_bench_verified,72.7,https://www.anthropic.com/news/claude-4,"{""effort"":""default"",""harness"":""simple bash+editor scaffold"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 10 trials)"",""temperature"":""top_p=0.95"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,Anthropic Claude 4 blog table: 72.7% single-pass / 80.2% w/ parallel.
claude-sonnet-4,mmlu,86.5,https://www.datacamp.com/blog/claude-4,{},,,dropped," [Dropped: DataCamp is a tutorial/learning platform aggregator, not primary source. Rule R5h-tutorial-blog.]"
claude-sonnet-4,tau_bench_retail,80.5,https://www.anthropic.com/news/claude-4,"{""effort"":""default"",""harness"":""\u03c4-bench official + max steps 100"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""prompt addendum to Agent Policy (Anthropic)"",""sampling"":""pass@1 (single-pass)"",""temperature"":""top_p=0.95"",""tools"":""agentic (TAU-bench official tools)""}",False,official_blog,verified,Anthropic Claude 4 blog table. Prompt addendum + max steps 100.
claude-sonnet-4,aime_2025,70.5,https://www.anthropic.com/news/claude-4,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (single-pass)"",""temperature"":""top_p=0.95"",""tools"":""none""}",True,official_blog,verified,Anthropic Claude 4 blog table: 70.5% single-pass / 85.0% w/ parallel. nucleus sampling top_p=0.95.
claude-sonnet-4,mmmu,74.4,https://www.anthropic.com/news/claude-4,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (single-pass)"",""temperature"":""top_p=0.95"",""tools"":""none""}",True,official_blog,verified,Anthropic Claude 4 blog table (MMMU validation).
claude-sonnet-4,swe_bench_pro,42.7,https://scale.com/leaderboard/swe_bench_pro_public,{},False,leaderboard,verified, [Verified against Scale.com SWE-Bench Pro Public leaderboard (2026-04-29): claude-4-Sonnet = 42.7. Scale uses standardized evaluation scaffolding.]
claude-sonnet-4,aime_2024,43.4,https://arxiv.org/abs/2507.20534,"{""context"":""8192 max output tokens"",""effort"":""default"",""harness"":""Kimi team evaluation (official API)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""avg@64"",""temperature"":""unified (Kimi team setting)"",""tools"":""none (benchmark-specified)""}",False,tech_report,verified,"Via third-party self-test by Moonshot AI (Kimi K2 tech report, Table 3, p.16). AIME2024 Avg@64, non-thinking mode (Kimi team eval), 8192 max output tokens."
claude-sonnet-4,arena_hard,51.6,https://arxiv.org/abs/2507.20534,"{""context"":""8192 max output tokens"",""effort"":""default"",""harness"":""Kimi team evaluation (official API)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (winrate)"",""temperature"":""unified (Kimi team setting)"",""tools"":""none (benchmark-specified)""}",False,tech_report,verified,"Via third-party self-test by Moonshot AI (Kimi K2 tech report, Table 3, p.16). Arena Hard v2.0 Hard Prompt winrate, non-thinking, 8192 max output tokens."
claude-sonnet-4,hle,5.8,https://arxiv.org/abs/2507.20534,"{""context"":""8192 max output tokens"",""effort"":""default"",""harness"":""Kimi team evaluation (official API)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unified (Kimi team setting)"",""tools"":""none (benchmark-specified)""}",False,tech_report,verified,"Via third-party self-test by Moonshot AI (Kimi K2 tech report, Table 3, p.16). HLE Text-Only Acc, non-thinking, 8192 max output tokens."
claude-sonnet-4,hmmt_feb_2025,15.9,https://arxiv.org/abs/2507.20534,"{""context"":""8192 max output tokens"",""effort"":""default"",""harness"":""Kimi team evaluation (official API)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""avg@32"",""temperature"":""unified (Kimi team setting)"",""tools"":""none (benchmark-specified)""}",False,tech_report,verified,"Via third-party self-test by Moonshot AI (Kimi K2 tech report, Table 3, p.16). HMMT 2025 Avg@32, non-thinking, 8192 max output tokens."
claude-sonnet-4,humaneval,88.0,https://www.datacamp.com/blog/claude-4,{},,,dropped," [Dropped: DataCamp is a tutorial/learning platform aggregator, not primary source. Rule R5h-tutorial-blog.]"
claude-sonnet-4,ifeval,87.6,https://arxiv.org/abs/2507.20534,"{""context"":""8192 max output tokens"",""effort"":""default"",""harness"":""Kimi team evaluation (official API)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unified (Kimi team setting)"",""tools"":""none (benchmark-specified)""}",False,tech_report,verified,"Via third-party self-test by Moonshot AI (Kimi K2 tech report, Table 3, p.16). IFEval Prompt Strict, non-thinking, 8192 max output tokens."
claude-sonnet-4,livecodebench,66.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2 model card: claude-sonnet-4=66.
claude-sonnet-4,math_500,94.0,https://arxiv.org/abs/2507.20534,"{""context"":""8192 max output tokens"",""effort"":""default"",""harness"":""Kimi team evaluation (official API)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unified (Kimi team setting)"",""tools"":""none (benchmark-specified)""}",False,tech_report,verified,"Via third-party self-test by Moonshot AI (Kimi K2 tech report, Table 3, p.16). MATH-500 Acc, non-thinking mode (Kimi team eval), 8192 max output tokens."
claude-sonnet-4,mmlu_pro,84.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2 model card: claude-sonnet-4=84.
claude-sonnet-4,simpleqa,15.9,https://arxiv.org/abs/2507.20534,"{""context"":""8192 max output tokens"",""effort"":""default"",""harness"":""Kimi team evaluation (official API)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unified (Kimi team setting)"",""tools"":""none (benchmark-specified)""}",False,tech_report,verified,"Via third-party self-test by Moonshot AI (Kimi K2 tech report, Table 3, p.16). SimpleQA Correct, non-thinking, 8192 max output tokens."
claude-sonnet-4,arc_agi_1,40.0,https://arcprize.org/leaderboard,"{""effort"":""16K thinking"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""CoT"",""tools"":""none""}",True,leaderboard,verified,arcprize.org leaderboard audit: Claude Sonnet 4 (Thinking 16K) on leaderboard
claude-sonnet-4,arc_agi_2,5.9,https://arcprize.org/leaderboard,"{""effort"":""16K thinking"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""CoT"",""tools"":""none""}",True,leaderboard,verified,arcprize.org leaderboard audit: Claude Sonnet 4 (Thinking 16K) on leaderboard
claude-sonnet-4,ifbench,55.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2 model card: claude-sonnet-4=55.
claude-sonnet-4,osworld,42.0,https://www.datacamp.com/blog/claude-4,{},,,dropped," [Dropped: DataCamp is a tutorial/learning platform aggregator, not primary source. Rule R5h-tutorial-blog.]"
claude-sonnet-4,terminal_bench,35.5,https://www.anthropic.com/news/claude-4,"{""effort"":""default"",""harness"":""Claude Code agent framework"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (single-pass)"",""temperature"":""top_p=0.95"",""tools"":""agentic""}",True,official_blog,verified,Anthropic Claude 4 blog table: 35.5% single-pass (Claude Code) / 41.3% w/ parallel. Same agent as non-Claude=33.5%.
claude-sonnet-4,terminal_bench_1,36.4,https://www.tbench.ai/leaderboard/terminal-bench/1.0,{},,leaderboard,verified,"Confirmed 36.4% on tbench.ai Terminal-Bench 1.0 leaderboard (Terminus 2 scaffold, 2025-08-05)."
claude-opus-4,gpqa_diamond,79.6,https://www.anthropic.com/news/claude-4,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (single-pass)"",""temperature"":""top_p=0.95"",""tools"":""none""}",True,official_blog,verified,Anthropic Claude 4 blog table: 79.6% single-pass / 83.3% w/ parallel test-time compute.
claude-opus-4,swe_bench_verified,72.5,https://www.anthropic.com/news/claude-4,"{""effort"":""default"",""harness"":""simple bash+editor scaffold"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 10 trials)"",""temperature"":""top_p=0.95"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,Anthropic Claude 4 blog table: 72.5% single-pass / 79.4% w/ parallel test-time compute. We record single-pass.
claude-opus-4,mmlu,88.8,https://www.datacamp.com/blog/claude-4,{},,,dropped," [Dropped: DataCamp is a tutorial/learning platform aggregator, not primary source. Rule R5h-tutorial-blog.]"
claude-opus-4,aime_2025,75.5,https://www.anthropic.com/news/claude-4,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (single-pass)"",""temperature"":""top_p=0.95"",""tools"":""none""}",True,official_blog,verified,Anthropic Claude 4 blog table: 75.5% single-pass / 90.0% w/ parallel test-time compute. nucleus sampling top_p=0.95.
claude-opus-4,tau_bench_retail,81.4,https://www.anthropic.com/news/claude-4,"{""effort"":""default"",""harness"":""\u03c4-bench official + max steps 100"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""prompt addendum to Agent Policy (Anthropic)"",""sampling"":""pass@1 (single-pass)"",""temperature"":""top_p=0.95"",""tools"":""agentic (TAU-bench official tools)""}",False,official_blog,verified,Anthropic Claude 4 blog table. With extended thinking + tool use; prompt addendum to Retail Agent Policy; max steps 100.
claude-opus-4,aime_2024,75.5,https://www.datacamp.com/blog/claude-4,{},,,dropped," [Dropped: DataCamp is a tutorial/learning platform aggregator, not primary source. Rule R5h-tutorial-blog.]"
claude-opus-4,mmmu,76.5,https://www.anthropic.com/news/claude-4,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (single-pass)"",""temperature"":""top_p=0.95"",""tools"":""none""}",True,official_blog,verified,Anthropic Claude 4 blog table (MMMU validation). With extended thinking up to 64K.
claude-opus-4,arena_hard,59.7,https://arxiv.org/abs/2507.20534,"{""context"":""8192 max output tokens"",""effort"":""default"",""harness"":""Kimi team evaluation (official API)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (winrate)"",""temperature"":""unified (Kimi team setting)"",""tools"":""none (benchmark-specified)""}",False,tech_report,verified,"Via third-party self-test by Moonshot AI (Kimi K2 tech report, Table 3, p.16). Arena Hard v2.0 Hard Prompt winrate, non-thinking, 8192 max output tokens."
claude-opus-4,frontiermath,4.5,https://epoch.ai/benchmarks/frontiermath,"{""effort"":""default"",""mode"":""thinking"",""tools"":""none""}",True,leaderboard,verified,"epoch.ai FrontierMath-2025-02-28-Private: claude-opus-4-20250514 (Claude Opus 4) = 4.5% (default thinking, no explicit budget). Canonical effort=default matches. BP had 10.0 (unverified)."
claude-opus-4,hle,7.1,https://arxiv.org/abs/2507.20534,"{""context"":""8192 max output tokens"",""effort"":""default"",""harness"":""Kimi team evaluation (official API)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unified (Kimi team setting)"",""tools"":""none (benchmark-specified)""}",False,tech_report,verified,"Via third-party self-test by Moonshot AI (Kimi K2 tech report, Table 3, p.16). HLE Text-Only Acc, non-thinking, 8192 max output tokens."
claude-opus-4,hmmt_feb_2025,15.9,https://arxiv.org/abs/2507.20534,"{""context"":""8192 max output tokens"",""effort"":""default"",""harness"":""Kimi team evaluation (official API)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""avg@32"",""temperature"":""unified (Kimi team setting)"",""tools"":""none (benchmark-specified)""}",False,tech_report,verified,"Via third-party self-test by Moonshot AI (Kimi K2 tech report, Table 3, p.16). HMMT 2025 Avg@32, non-thinking, 8192 max output tokens."
claude-opus-4,ifeval,89.7,https://huggingface.co/Qwen/Qwen3-235B-A22B-Thinking-2507,{},True,model_card,verified,
claude-opus-4,livecodebench,47.4,https://arxiv.org/abs/2507.20534,"{""context"":""8192 max output tokens"",""effort"":""default"",""harness"":""Kimi team evaluation (official API)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unified (Kimi team setting)"",""tools"":""none (benchmark-specified)""}",False,tech_report,verified,"Via third-party self-test by Moonshot AI (Kimi K2 tech report, Table 3, p.16). LiveCodeBench v6 (Aug 2024–May 2025) Pass@1, non-thinking, 8192 max output tokens."
claude-opus-4,math_500,94.4,https://arxiv.org/abs/2507.20534,"{""context"":""8192 max output tokens"",""effort"":""default"",""harness"":""Kimi team evaluation (official API)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unified (Kimi team setting)"",""tools"":""none (benchmark-specified)""}",False,tech_report,verified,"Via third-party self-test by Moonshot AI (Kimi K2 tech report, Table 3, p.16). MATH-500 Acc, non-thinking mode (Kimi team eval), 8192 max output tokens."
claude-opus-4,mmlu_pro,86.6,https://arxiv.org/abs/2507.20534,"{""context"":""8192 max output tokens"",""effort"":""default"",""harness"":""Kimi team evaluation (official API)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (EM)"",""temperature"":""unified (Kimi team setting)"",""tools"":""none (benchmark-specified)""}",False,tech_report,verified,"Via third-party self-test by Moonshot AI (Kimi K2 tech report, Table 3, p.16). MMLU-Pro EM, non-thinking, 8192 max output tokens."
claude-opus-4,simpleqa,22.8,https://arxiv.org/abs/2507.20534,"{""context"":""8192 max output tokens"",""effort"":""default"",""harness"":""Kimi team evaluation (official API)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unified (Kimi team setting)"",""tools"":""none (benchmark-specified)""}",False,tech_report,verified,"Via third-party self-test by Moonshot AI (Kimi K2 tech report, Table 3, p.16). SimpleQA Correct, non-thinking, 8192 max output tokens."
claude-opus-4,arc_agi_1,35.7,https://arcprize.org/leaderboard,"{""effort"":""16K thinking"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""CoT"",""tools"":""none""}",True,leaderboard,verified,arcprize.org leaderboard audit: Claude Opus 4 (Thinking 16K) on leaderboard
claude-opus-4,arc_agi_2,8.6,https://arcprize.org/leaderboard,"{""effort"":""16K thinking"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""CoT"",""tools"":""none""}",True,leaderboard,verified,arcprize.org leaderboard audit: Claude Opus 4 (Thinking 16K) on leaderboard
claude-opus-4,codeforces_rating,1886.0,https://www.datacamp.com/blog/claude-4,{},,,dropped," [Dropped: DataCamp is a tutorial/learning platform aggregator, not primary source. Rule R5h-tutorial-blog.]"
claude-opus-4,critpt,0.3,https://github.com/CritPt-Benchmark/CritPt,"""CritPt leaderboard, \u03b5=1.5, % tasks solved""",True,leaderboard,verified,
claude-opus-4,osworld,38.2,https://www.datacamp.com/blog/claude-4,{},,,dropped," [Dropped: DataCamp is a tutorial/learning platform aggregator, not primary source. Rule R5h-tutorial-blog.]"
claude-opus-4,swe_bench_pro,35.8,https://www.anthropic.com/news/claude-opus-4-1,{},,,dropped,
claude-opus-4,terminal_bench,43.2,https://www.anthropic.com/news/claude-4,"{""effort"":""default"",""harness"":""Claude Code agent framework"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (single-pass)"",""temperature"":""top_p=0.95"",""tools"":""agentic""}",True,official_blog,verified,Anthropic Claude 4 blog table: 43.2% single-pass (Claude Code agent) / 50.0% w/ parallel. Same agent as non-Claude=39.2%.
claude-opus-4,terminal_bench_1,39.0,https://www.tbench.ai/leaderboard/terminal-bench/1.0,{},,leaderboard,verified,"Confirmed 39.0% on tbench.ai Terminal-Bench 1.0 leaderboard (Terminus 2 scaffold, 2025-08-05). score=39.0"
claude-opus-4.1,swe_bench_verified,74.5,https://www.anthropic.com/news/claude-opus-4-1,"{""effort"":""default"",""harness"":""simple bash+editor scaffold"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 10 trials)"",""temperature"":""top_p=0.95"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,"Anthropic Opus 4.1 blog table: 74.5%, single-pass."
claude-opus-4.1,gpqa_diamond,80.9,https://www.anthropic.com/news/claude-opus-4-1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (single-pass)"",""temperature"":""top_p=0.95"",""tools"":""none""}",True,official_blog,verified,Anthropic Opus 4.1 blog table: 80.9% (with extended thinking).
claude-opus-4.1,mmlu,88.8,https://www.anthropic.com/news/claude-opus-4-1,{},,,needs_review,Anthropic blog benchmark table is a CDN PNG image; cannot extract text value. R5g(c): value not found in source text.
claude-opus-4.1,mmlu_pro,87.92,https://artificialanalysis.ai/evaluations/mmlu-pro,{},,,needs_review,"Reference URL is AA mmlu-pro eval page. AA shows claude-4-1-opus-thinking mmlu_pro=0.88→88.0%, but BP has 87.92. Close but not exact match; AA may have rounded. Only thinking variant available on AA. Current AA value: 88.0%."
claude-opus-4.1,swe_bench_pro,22.7,https://scale.com/leaderboard/swe_bench_pro_public,{},,,verified,"Confirmed 22.7% from scale.com analysis text: ""Claude Opus 4.1 decreases from 22.7% to 17.8% resolution""; historical public score."
claude-opus-4.1,aime_2025,78.0,https://www.anthropic.com/news/claude-opus-4-1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (single-pass)"",""temperature"":""top_p=0.95"",""tools"":""none""}",True,official_blog,verified,"Anthropic Opus 4.1 blog table: 78.0%, with extended thinking, top_p=0.95."
claude-opus-4.1,frontiermath,5.9,https://epoch.ai/benchmarks/frontiermath,"{""effort"":""default"",""mode"":""thinking"",""tools"":""none""}",True,leaderboard,verified,epoch.ai FrontierMath-2025-02-28-Private: claude-opus-4-1-20250805 (no explicit thinking budget) = 5.9%. Canonical effort=default matches. BP had 15.0.
claude-opus-4.1,osworld,44.4,https://assets.anthropic.com/m/64823ba7485345a7/Claude-Opus-4-5-System-Card.pdf,{},,,verified,FLAGGED cleared: OSWorld=44.4% confirmed in cross-model Table 2.3.A (p.19) of Opus 4.5 system card.
claude-opus-4.1,terminal_bench,43.3,https://www.anthropic.com/news/claude-opus-4-1,"{""effort"":""default"",""harness"":""Terminus 1"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""top_p=0.95"",""tools"":""agentic""}",True,official_blog,verified,"Anthropic Opus 4.1 blog table: 43.3%, Terminus 1, avg 5 trials. Differs from Opus 4 (Claude Code framework=43.2)."
claude-opus-4.1,hle,35.0,https://www.anthropic.com/news/claude-opus-4-1,{},,,flagged,Anthropic blog benchmark table is CDN PNG; cannot confirm or deny flag. Keeping flagged.
claude-opus-4.1,humaneval,93.0,https://www.anthropic.com/news/claude-opus-4-1,{},,,flagged,Anthropic blog benchmark table is CDN PNG; cannot confirm or deny flag. Keeping flagged.
claude-opus-4.1,livebench,61.4,https://livebench.ai/,"{""context"":""32k thinking"",""effort"":""default"",""mode"":""thinking"",""tools"":""n/a""}",True,leaderboard,verified,livebench.ai table_2026_01_08 (latest as of 2026-04-28): claude-4-1-opus-20250805-thinking-32k avg=61.4. Canonical thinking mode matches. BP had 54.5.
claude-opus-4.1,livecodebench,63.2,https://www.anthropic.com/news/claude-opus-4-1,{},,,flagged,Anthropic blog benchmark table is CDN PNG; cannot confirm or deny flag. Keeping flagged.
claude-opus-4.1,simpleqa,43.5,https://www.anthropic.com/news/claude-opus-4-1,{},,,flagged,Anthropic blog benchmark table is CDN PNG; cannot confirm or deny flag. Keeping flagged.
claude-opus-4.1,terminal_bench_1,43.8,https://www.tbench.ai/leaderboard/terminal-bench/1.0,{},,,verified,"Confirmed 43.8% on tbench.ai Terminal-Bench 1.0 leaderboard (Terminus 2 scaffold, 2025-09-30)."
claude-sonnet-4.5,swe_bench_verified,77.2,https://www.anthropic.com/news/claude-opus-4-5,"{""effort"":""high"",""harness"":""simple bash+editor scaffold"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""interleaved scratchpads"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,Anthropic Opus 4.5 blog table: Sonnet 4.5 = 77.2% (no thinking).
claude-sonnet-4.5,gpqa_diamond,83.4,https://www.anthropic.com/news/claude-opus-4-5,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""interleaved scratchpads"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,Anthropic Opus 4.5 blog table: Sonnet 4.5 = 83.4%.
claude-sonnet-4.5,tau_bench_retail,86.2,https://assets.anthropic.com/m/64823ba7485345a7/Claude-Opus-4-5-System-Card.pdf,{},,,verified,Confirmed τ²-Bench Retail=86.2% in cross-model Table 2.3.A (p.19) of Opus 4.5 system card.
claude-sonnet-4.5,tau_bench_telecom,98.0,https://assets.anthropic.com/m/64823ba7485345a7/Claude-Opus-4-5-System-Card.pdf,{},,,verified,Confirmed τ²-Bench Telecom=98.0% in cross-model Table 2.3.A (p.19) of Opus 4.5 system card.
claude-sonnet-4.5,osworld,61.4,https://www.anthropic.com/news/claude-opus-4-5,"{""effort"":""high"",""harness"":""OSWorld official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""interleaved scratchpads"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic (browser/computer use)""}",True,official_blog,verified,Anthropic Opus 4.5 blog table: Sonnet 4.5 = 61.4%.
claude-sonnet-4.5,livecodebench,64.0,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: claude-sonnet-4.5=64.0.
claude-sonnet-4.5,aime_2025,87.0,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"DeepMind /models/gemini/flash/ Performance table: AIME 2025 no tools 87.0% [Promoted to verified, prior unverified value deleted per R5d.]"
claude-sonnet-4.5,humaneval,85.0,https://www.getpassionfruit.com/blog/gpt-5-1-vs-claude-4-5-sonnet-vs-gemini-3-pro-vs-deepseek-v3-2-the-definitive-2025-ai-model-comparison,{},,,needs_review,Source is third-party blog (getpassionfruit.com); R5h priority. Not verifiable from official Anthropic source.
claude-sonnet-4.5,mmmu,77.8,https://www.anthropic.com/news/claude-opus-4-5,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""interleaved scratchpads"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,Anthropic Opus 4.5 blog table: Sonnet 4.5 = 77.8%.
claude-sonnet-4.5,mmmu_pro,63.4,https://www.anthropic.com/news/claude-sonnet-4-6,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""updated grading: separate Sonnet 4 grader; no 'think step-by-step' prefix"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,Anthropic Sonnet 4.6 blog table: claude-sonnet-4.5 = 63.4.
claude-sonnet-4.5,swe_bench_pro,48.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-sonnet-4.5=48.4.
claude-sonnet-4.5,mrcr_v2,47.1,https://deepmind.google/models/gemini/flash/,"{""context"":""128k"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"DeepMind /models/gemini/flash/ Performance table: MRCR v2 8-needle 128k 47.1% [Promoted to verified, prior unverified value deleted per R5d.]"
claude-sonnet-4.5,mmlu,86.5,https://www.leanware.co/insights/claude-sonnet-4-5-overview,{},,,needs_review,Source is third-party blog (leanware.co); R5h priority. Official source has image-only table.
claude-sonnet-4.5,codeforces_rating,1480.0,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: claude-sonnet-4.5=1480.
claude-sonnet-4.5,hle,17.7,https://www.anthropic.com/news/claude-sonnet-4-6,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none (no-tools column)""}",True,official_blog,verified,Anthropic Sonnet 4.6 blog table: claude-sonnet-4.5 = 17.7.
claude-sonnet-4.5,mmlu_pro,88.2,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: claude-sonnet-4.5=88.2.
claude-sonnet-4.5,terminal_bench,50.0,https://www.anthropic.com/news/claude-opus-4-5,"{""effort"":""128K thinking"",""harness"":""Terminus-2"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""interleaved scratchpads"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,"Anthropic Opus 4.5 blog table: Sonnet 4.5 = 50.0%, Terminus-2 harness."
claude-sonnet-4.5,aime_2024,88.0,https://www.leanware.co/insights/claude-sonnet-4-5-overview,{},,,needs_review,Source is third-party blog (leanware.co); R5h priority.
claude-sonnet-4.5,arc_agi_1,70.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-sonnet-4.5=70.9.
claude-sonnet-4.5,arc_agi_2,13.6,https://www.anthropic.com/news/claude-opus-4-5,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""interleaved scratchpads"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,Anthropic Opus 4.5 blog table: Sonnet 4.5 = 13.6%.
claude-sonnet-4.5,brumo_2025,90.83,https://matharena.ai/?comp=brumo--brumo_2025,{},,,verified,Confirmed brumo_2025=90.83% on matharena.ai model page.
claude-sonnet-4.5,cmimc_2025,66.88,https://matharena.ai/?comp=cmimc--cmimc_2025,{},,,verified,Confirmed cmimc_2025=66.88% on matharena.ai model page.
claude-sonnet-4.5,ifeval,88.5,https://www.leanware.co/insights/claude-sonnet-4-5-overview,{},,,needs_review,Source is third-party blog (leanware.co); R5h priority.
claude-sonnet-4.5,livebench,67.9,https://livebench.ai/,"{""context"":""64k thinking"",""effort"":""high"",""mode"":""thinking"",""tools"":""n/a""}",True,leaderboard,verified,livebench.ai table_2026_01_08 (latest as of 2026-04-28): claude-sonnet-4-5-20250929-thinking-64k avg=67.9. Canonical thinking+high effort matches. BP had 53.7.
claude-sonnet-4.5,math_500,95.8,https://www.leanware.co/insights/claude-sonnet-4-5-overview,{},,,needs_review,Source is third-party blog (leanware.co); R5h priority.
claude-sonnet-4.5,matharena_apex_2025,1.56,https://matharena.ai/?comp=apex--apex_2025,{},,,verified,Confirmed matharena_apex_2025=1.56% on matharena.ai model page.
claude-sonnet-4.5,simpleqa,47.0,https://www.vellum.ai/blog/claude-opus-4-5-benchmarks,{},,,dropped,"Source is third-party blog (vellum.ai); R5h priority. DROPPED (R5h): random third-party blog/aggregator, no primary trail"
claude-sonnet-4.5,smt_2025,83.96,https://matharena.ai/?comp=smt--smt_2025,{},,,verified,Confirmed smt_2025=83.96% on matharena.ai model page.
claude-sonnet-4.5,terminal_bench_1,51.0,https://www.tbench.ai/leaderboard/terminal-bench/1.0,{},,,verified,"Confirmed 51.0% on tbench.ai Terminal-Bench 1.0 leaderboard (Terminus 2 scaffold, 2025-10-31)."
claude-haiku-4.5,mmmlu,83.0,https://www.anthropic.com/news/claude-haiku-4-5,"{""effort"":""128K thinking budget"",""harness"":""official"",""judge"":""official harness"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"Anthropic blog, avg 10 runs over 14 non-English languages, 128K thinking budget."
claude-haiku-4.5,mmmu,73.2,https://www.anthropic.com/news/claude-haiku-4-5,"{""effort"":""128K thinking budget"",""harness"":""official"",""judge"":""official harness"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"Anthropic blog (MMMU validation), 128K thinking, avg 10 runs."
claude-haiku-4.5,swe_bench_verified,73.3,https://www.anthropic.com/news/claude-haiku-4-5,"{""effort"":""128K thinking budget"",""harness"":""official"",""judge"":""official harness"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + file editing)""}",False,official_blog,verified,"Anthropic blog, full 500 instances, avg 50 trials, 128K thinking, default sampling. NOTE: Anthropic added prompt addendum (""use tools 100+ times, implement own tests first"") — deviates from default prompt → matches_canonical=false."
claude-haiku-4.5,gpqa_diamond,73.0,https://www.anthropic.com/news/claude-haiku-4-5,"{""effort"":""128K thinking budget"",""harness"":""official"",""judge"":""official harness"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"Anthropic blog, 128K thinking, avg over 10 runs. Prior BP value 80.9 not found on blog (corrected)."
claude-haiku-4.5,aime_2025,96.3,https://www.anthropic.com/news/claude-haiku-4-5,"{""effort"":""128K thinking budget"",""harness"":""official"",""judge"":""official harness"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""python interpreter""}",False,official_blog,verified,"Anthropic blog, with python tool, avg 10 runs of pass@1 over 16 trials, 128K thinking. Blog also reports 80.7% no-tools (canonical-matching). Updated source from airank.dev to Anthropic primary."
claude-haiku-4.5,osworld,50.7,https://www.anthropic.com/news/claude-haiku-4-5,"{""effort"":""128K thinking budget"",""harness"":""official"",""judge"":""official harness"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (OSWorld-Verified, 100 max steps)""}",True,official_blog,verified,"Anthropic blog, official OSWorld-Verified framework, avg 4 runs, 128K total + 2K per-step thinking budget."
claude-haiku-4.5,swe_bench_pro,39.45,https://scale.com/leaderboard/swe_bench_pro_public,{},,,verified,Confirmed 39.45% on scale.com SWE-bench Pro Public leaderboard (current live entry).
claude-haiku-4.5,terminal_bench,41.0,https://www.anthropic.com/news/claude-haiku-4-5,"{""effort"":""128K thinking budget"",""harness"":""official"",""judge"":""official harness"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (Terminus 2 default scaffold)""}",True,official_blog,verified,"Anthropic blog, Terminus 2 default agent, avg 11 runs (6 no-thinking 40.21%, 5 with 32K thinking 41.75%). Replaces prior tbench.ai leaderboard value 28.3."
claude-haiku-4.5,arc_agi_1,47.7,https://arcprize.org/leaderboard,"{""effort"":""32K thinking"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""CoT"",""tools"":""none""}",False,leaderboard,verified,arcprize.org leaderboard audit: Claude Haiku 4.5 (Thinking 32K) on leaderboard; canonical is 128K
claude-haiku-4.5,arc_agi_2,4.0,https://arcprize.org/leaderboard,"{""effort"":""32K thinking"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""CoT"",""tools"":""none""}",False,leaderboard,verified,arcprize.org leaderboard audit: Claude Haiku 4.5 (Thinking 32K) on leaderboard; canonical is 128K
claude-haiku-4.5,humaneval,91.5,https://www.anthropic.com/claude/haiku,{},,,needs_review,Anthropic claude/haiku model page benchmark table is CDN PNG image; cannot extract text value. R5g(c).
claude-haiku-4.5,ifeval,85.0,https://www.anthropic.com/claude/haiku,{},,,needs_review,Anthropic claude/haiku model page benchmark table is CDN PNG image; cannot extract text value. R5g(c).
claude-haiku-4.5,livebench,61.0,https://livebench.ai/,"{""context"":""64k thinking"",""effort"":""n/a"",""mode"":""thinking"",""tools"":""n/a""}",False,leaderboard,verified,livebench.ai table_2026_01_08 (latest as of 2026-04-28): claude-haiku-4-5-20251001-thinking-64k avg=61.0. Canonical requires 128K thinking budget but livebench only has 64k variant; used 64k as best available. BP had 45.3.
claude-haiku-4.5,livecodebench,52.0,https://artificialanalysis.ai/models/claude-4-5-haiku,{},,,needs_review,"AA model page (claude-4-5-haiku): non-reasoning variant livecodebench=0.511→51.1%; reasoning variant=0.615→61.5%. BP has 52, which matches neither. BP canonical is thinking mode but neither AA variant shows 52. Current AA (non-reasoning): 51.1%, (reasoning): 61.5%."
claude-haiku-4.5,math_500,90.2,https://www.anthropic.com/claude/haiku,{},,,needs_review,Anthropic claude/haiku model page benchmark table is CDN PNG image; cannot extract text value. R5g(c).
claude-haiku-4.5,mmlu_pro,82.16,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 82.16
claude-opus-4.5,swe_bench_verified,80.9,https://www.anthropic.com/news/claude-opus-4-5,"{""effort"":""high"",""harness"":""simple bash+editor scaffold"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""interleaved scratchpads"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,Anthropic Opus 4.5 blog table: 80.9% (no thinking). avg 5 trials.
claude-opus-4.5,gpqa_diamond,87.0,https://www.anthropic.com/news/claude-opus-4-5,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""interleaved scratchpads"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,Anthropic Opus 4.5 blog table: 87.0% (with extended thinking 64K).
claude-opus-4.5,mmlu,90.8,https://assets.anthropic.com/m/64823ba7485345a7/Claude-Opus-4-5-System-Card.pdf,{},,,verified,MMMLU=90.77 in Table 2.3.A (p.19) of Opus 4.5 system card; rounds to 90.8 as reported.
claude-opus-4.5,mmlu_pro,89.3,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: claude-opus-4.5=89.3.
claude-opus-4.5,osworld,66.3,https://www.anthropic.com/news/claude-opus-4-5,"{""effort"":""high"",""harness"":""OSWorld official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""interleaved scratchpads"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic (browser/computer use)""}",True,official_blog,verified,Anthropic Opus 4.5 blog table: 66.3% (computer use).
claude-opus-4.5,arc_agi_2,37.6,https://www.anthropic.com/news/claude-opus-4-5,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""interleaved scratchpads"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"Anthropic Opus 4.5 blog table: 37.6% (Verified, novel problem solving)."
claude-opus-4.5,browsecomp,67.8,https://www.anthropic.com/news/claude-sonnet-4-6,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic (web search + fetch + programmatic tool calling, context compaction at 50k up to 10M)""}",True,official_blog,verified,Anthropic Sonnet 4.6 blog table: claude-opus-4.5 = 67.8.
claude-opus-4.5,frontiermath,20.7,https://epoch.ai/benchmarks/frontiermath,"{""context"":""32k thinking"",""effort"":""high"",""mode"":""thinking"",""tools"":""none""}",False,leaderboard,verified,epoch.ai FrontierMath-2025-02-28-Private: claude-opus-4-5-20251101_32K (32k thinking) = 20.7%. Canonical requires 64K thinking budget but epoch only has 16k (20.3%) and 32k (20.7%); used 32k as closest. BP had 21.0 (likely rounded from 20.7).
claude-opus-4.5,hle,30.8,https://www.anthropic.com/news/claude-sonnet-4-6,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none (no-tools column)""}",True,official_blog,verified,Anthropic Sonnet 4.6 blog table: claude-opus-4.5 = 30.8.
claude-opus-4.5,tau_bench_telecom,98.2,https://assets.anthropic.com/m/64823ba7485345a7/Claude-Opus-4-5-System-Card.pdf,{},,,verified,Confirmed τ²-Bench Telecom=98.2% in Table 2.3.A (p.19) of Opus 4.5 system card.
claude-opus-4.5,tau_bench_retail,88.9,https://assets.anthropic.com/m/64823ba7485345a7/Claude-Opus-4-5-System-Card.pdf,{},,,verified,Confirmed τ²-Bench Retail=88.9% in Table 2.3.A (p.19) of Opus 4.5 system card.
claude-opus-4.5,swe_bench_pro,55.4,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic (bash + edit)""}",True,model_card,verified,Kimi K2.5 model card: claude-opus-4.5=55.4.
claude-opus-4.5,aime_2025,92.8,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: claude-opus-4.5=92.8.
claude-opus-4.5,simpleqa,72.0,https://artificialanalysis.ai/articles/claude-opus-4-5-benchmarks-and-analysis,{},,,needs_review,Reference URL is AA article page (claude-opus-4-5-benchmarks-and-analysis). AA does not publish SimpleQA scores; field is not present in AA model data for any model. Cannot verify from AA source. Original source needs identification.
claude-opus-4.5,chatbot_arena_elo,1469.0,https://lmarena.ai/leaderboard/text,{},True,leaderboard,verified,lmarena.ai text arena (fetched 2026-04-29). Arena model name: claude-opus-4-5-20251101 (rank 19). ELO updates continuously; score reflects latest available.
claude-opus-4.5,mmmu,80.7,https://www.anthropic.com/news/claude-opus-4-5,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""interleaved scratchpads"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"Anthropic Opus 4.5 blog table: 80.7% (validation, with extended thinking)."
claude-opus-4.5,math_500,85.0,https://www.vellum.ai/blog/claude-opus-4-5-benchmarks,{},,,dropped,"Source is third-party blog (vellum.ai); R5h priority. DROPPED (R5h): random third-party blog/aggregator, no primary trail"
claude-opus-4.5,ifeval,90.9,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,model_card,verified,"Qwen3.5-397B-A17B HF model card cross-model table (Qwen self-test of 3rd-party model, mc=false): ifeval 90.9 (3rd-party Qwen self-test). [R5d: prior unverified value 90.0 from https://www.vellum.ai/blog/claude-opus-4-5-benchmarks deleted.]"
claude-opus-4.5,livecodebench,84.8,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,model_card,verified,"Qwen3.5-397B-A17B HF model card cross-model table (Qwen self-test of 3rd-party model, mc=false): livecodebench 84.8 (3rd-party Qwen self-test). [R5d: prior unverified value 68.0 from https://www.anthropic.com/news/claude-opus-4-5 deleted.]"
claude-opus-4.5,video_mmmu,84.4,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: claude-opus-4.5=84.4.
claude-opus-4.5,aa_lcr,71.3,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: claude-opus-4.5=71.3.
claude-opus-4.5,terminal_bench,59.3,https://www.anthropic.com/news/claude-opus-4-5,"{""effort"":""128K thinking"",""harness"":""Terminus-2"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""interleaved scratchpads"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,"Anthropic Opus 4.5 blog table: 59.3%, Terminus-2 harness, 128K thinking, avg 5 trials."
claude-opus-4.5,aime_2024,90.0,https://artificialanalysis.ai/articles/claude-opus-4-5-benchmarks-and-analysis,{},,,needs_review,Reference URL is AA article page. Current AA data: claude-4-opus-thinking (reasoning) aime=0.757→75.7%; claude-4-opus (non-thinking) aime=0.563→56.3%. BP has 90 which matches neither. 90% may be Anthropic lab-claimed value mistakenly attributed to AA. Current AA (thinking): 75.7%.
claude-opus-4.5,arc_agi_1,84.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-opus-4.5=84.0.
claude-opus-4.5,codeforces_rating,1701.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-opus-4.5=1701.
claude-opus-4.5,humaneval,95.1,https://artificialanalysis.ai/articles/claude-opus-4-5-benchmarks-and-analysis,{},,,needs_review,"Reference URL is AA article page. Current AA data: claude-4-opus (non-thinking) humaneval=0.970→97.0%, claude-4-opus-thinking=N/A. BP has 95.1% which does not match AA current value (97.0%). Possible stale/different evaluation config."
claude-opus-4.5,livebench,76.0,https://livebench.ai/,"{""context"":""64k thinking"",""effort"":""high"",""mode"":""thinking"",""tools"":""n/a""}",True,leaderboard,verified,livebench.ai table_2026_01_08 (latest as of 2026-04-28): claude-opus-4-5-20251101-thinking-64k-high-effort avg=76.0. EXACT MATCH with BP value 76. Canonical thinking+high+64K matches.
claude-opus-4.5,simplebench,62.0,https://lmcouncil.ai/benchmarks,{},,,dropped,"lmcouncil.ai scores are JS-rendered; could not extract value. R5g(c). DROPPED (R5h): random third-party blog/aggregator, no primary trail | DROPPED (R5g-c stale, never verified, source not audited in any pass)"
claude-opus-4.6,gpqa_diamond,91.3,https://www-cdn.anthropic.com/8b8380204f74670be75e81c820ca8dda846ab289.pdf,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,"Mythos sys card Table 6.3.A: Opus 4.6 = 91.3%. Verified per Anthropic standard config (thinking max effort, default sampling, avg 5 trials)."
claude-opus-4.6,swe_bench_verified,80.8,https://www-cdn.anthropic.com/8b8380204f74670be75e81c820ca8dda846ab289.pdf,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic""}",True,tech_report,verified,"Mythos sys card Table 6.3.A: Opus 4.6 = 80.8%. Verified per Anthropic standard config (thinking max effort, default sampling, avg 5 trials)."
claude-opus-4.6,aime_2025,95.6,https://huggingface.co/MiniMaxAI/MiniMax-M2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2.5 model card: claude-opus-4.6=95.6.
claude-opus-4.6,mmlu_pro,89.1,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): claude-opus-4.6 (per their own canonical config) = 89.1.
claude-opus-4.6,osworld,72.7,https://www-cdn.anthropic.com/8b8380204f74670be75e81c820ca8dda846ab289.pdf,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic""}",True,tech_report,verified,"Mythos sys card Table 6.3.A: Opus 4.6 = 72.7%. Verified per Anthropic standard config (thinking max effort, default sampling, avg 5 trials)."
claude-opus-4.6,arc_agi_2,68.8,https://www.anthropic.com/news/claude-sonnet-4-6,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""120k thinking budget"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,Anthropic Sonnet 4.6 blog table: claude-opus-4.6 = 68.8.
claude-opus-4.6,browsecomp,83.7,https://www.anthropic.com/news/claude-opus-4-7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,Anthropic Opus 4.7 blog table: claude-opus-4.6 = 83.7%.
claude-opus-4.6,hle,40.0,https://www-cdn.anthropic.com/8b8380204f74670be75e81c820ca8dda846ab289.pdf,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,"Mythos sys card Table 6.3.A: Opus 4.6 = 40.0%. Verified per Anthropic standard config (thinking max effort, default sampling, avg 5 trials)."
claude-opus-4.6,frontiermath,40.7,https://epoch.ai/benchmarks/frontiermath,"{""effort"":""max"",""mode"":""thinking"",""tools"":""none""}",True,leaderboard,verified,"epoch.ai FrontierMath-2025-02-28-Private: claude-opus-4-6_max = 40.7% (adaptive thinking, max effort). Canonical effort=max matches. BP had 40.0 (matched 32k thinking variant, not max)."
claude-opus-4.6,mrcr_v2,84.0,https://deepmind.google/models/gemini/pro/,"{""context"":""128k"",""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/pro/ Performance table (PROMOTED from candidate; prior unverified value 93.0 from https://www.vellum.ai/blog/claude-opus-4-6-benchmarks deleted per audit rule). DeepMind /models/gemini/pro/ cross-model Performance table: MRCR v2 8-needle 128k average 84.0%
claude-opus-4.6,simpleqa,72.0,https://www.vellum.ai/blog/claude-opus-4-6-benchmarks,{},,,dropped,"Source is third-party blog (vellum.ai); R5h priority. DROPPED (R5h): random third-party blog/aggregator, no primary trail"
claude-opus-4.6,ifeval,92.2,https://longcat.chat/blog/longcat-2.0/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""Meituan unified in-house harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Measured in-house by Meituan unless otherwise noted."",""prompt_style"":""default"",""sampling"":""source does not state"",""temperature"":""0.7; top_k=-1; top_p=0.95"",""tools"":""none""}",False,official_blog,verified,LongCat-2.0 official tech blog: IFEval = 92.2. Measured in-house by Meituan under the reported unified harness.
claude-opus-4.6,humaneval,95.0,https://www.vellum.ai/blog/claude-opus-4-6-benchmarks,{},,,dropped,"Source is third-party blog (vellum.ai); R5h priority. DROPPED (R5h): random third-party blog/aggregator, no primary trail"
claude-opus-4.6,livecodebench,88.8,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): claude-opus-4.6 (per their own canonical config) = 88.8.
claude-opus-4.6,tau_bench_retail,91.9,https://www.vellum.ai/blog/claude-opus-4-6-benchmarks,{},,,dropped,"Source is third-party blog (vellum.ai); R5h priority. DROPPED (R5h): random third-party blog/aggregator, no primary trail"
claude-opus-4.6,critpt,12.6,https://artificialanalysis.ai/evaluations/critpt,"{""effort"":""max (adaptive reasoning)"",""harness"":""AA standard evaluation"",""mode"":""thinking"",""notes"":""AA slug: claude-opus-4-6-adaptive; critpt=12.6%"",""sampling"":""pass@1""}",,third_party,verified,ArtificialAnalysis aggregator; used as fallback when no primary source available (R5d-aggregator-fallback). AA CritPT eval page: claude-opus-4-6-adaptive critpt=0.126 → 12.6%. AA adaptive = max-effort adaptive reasoning variant.
claude-opus-4.6,gdpval_aa_elo,1606.0,https://www.anthropic.com/news/claude-sonnet-4-6,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,Anthropic Sonnet 4.6 blog table: claude-opus-4.6 = 1606.
claude-opus-4.6,aa_intelligence_index,53.0,https://artificialanalysis.ai/evaluations/artificial-analysis-intelligence-index,"{""effort"":""max (adaptive reasoning)"",""harness"":""AA Intelligence Index composite"",""mode"":""thinking"",""notes"":""AA slug: claude-opus-4-6-adaptive; II=52.95 rounds to 53""}",,third_party,verified,ArtificialAnalysis aggregator; used as fallback when no primary source available (R5d-aggregator-fallback). AA Intelligence Index: claude-opus-4-6-adaptive = 52.95 (displayed as 53). AA adaptive = max-effort adaptive reasoning variant.
claude-opus-4.6,mmmu,77.0,https://www.vellum.ai/blog/claude-opus-4-6-benchmarks,{},,,dropped,"Source is third-party blog (vellum.ai); R5h priority. DROPPED (R5h): random third-party blog/aggregator, no primary trail"
claude-opus-4.6,math_500,93.0,https://www.vellum.ai/blog/claude-opus-4-6-benchmarks,{},,,dropped,"Source is third-party blog (vellum.ai); R5h priority. DROPPED (R5h): random third-party blog/aggregator, no primary trail"
claude-opus-4.6,mmlu,90.8,https://www.vellum.ai/blog/claude-opus-4-6-benchmarks,{},,,dropped,"Source is third-party blog (vellum.ai); R5h priority. DROPPED (R5h): random third-party blog/aggregator, no primary trail"
claude-opus-4.6,chatbot_arena_elo,1496.0,https://lmarena.ai/leaderboard/text,{},True,leaderboard,verified,lmarena.ai text arena (fetched 2026-04-29). Arena model name: claude-opus-4-6 (rank 3). ELO updates continuously; score reflects latest available.
claude-opus-4.6,tau_bench_telecom,98.2,https://www.anthropic.com/news/claude-opus-4-6,{},,,needs_review,Anthropic blog benchmark table is a CDN PNG image; cannot extract text value. R5g(c). Note: Opus 4.5 system card shows tau_bench_telecom=98.2% for Opus 4.5; Opus 4.6 value is unverified.
claude-opus-4.6,swe_bench_pro,53.4,https://www-cdn.anthropic.com/8b8380204f74670be75e81c820ca8dda846ab289.pdf,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic""}",True,tech_report,verified,"Mythos sys card Table 6.3.A: Opus 4.6 = 53.4%. Verified per Anthropic standard config (thinking max effort, default sampling, avg 5 trials)."
claude-opus-4.6,arc_agi_1,93.0,https://arcprize.org/leaderboard,"{""effort"":""max (120K)"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""CoT"",""tools"":""none""}",True,leaderboard,verified,"arcprize.org leaderboard audit: Claude Opus 4.6 (120K, Max) on leaderboard"
claude-opus-4.6,codeforces_rating,2650.0,https://www.vellum.ai/blog/claude-opus-4-6-benchmarks,{},,,dropped,"Source is third-party blog (vellum.ai); R5h priority. DROPPED (R5h): random third-party blog/aggregator, no primary trail"
claude-opus-4.6,mmmu_pro,73.9,https://www.anthropic.com/news/claude-sonnet-4-6,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""updated grading: separate Sonnet 4 grader; no 'think step-by-step' prefix"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,Anthropic Sonnet 4.6 blog table: claude-opus-4.6 = 73.9.
claude-opus-4.6,terminal_bench,65.4,https://www-cdn.anthropic.com/8b8380204f74670be75e81c820ca8dda846ab289.pdf,"{""effort"":""max"",""harness"":""Harbor scaffold + Terminus-2"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic""}",True,tech_report,verified,"Mythos sys card Table 6.3.A: Opus 4.6 = 65.4%. Verified per Anthropic standard config (thinking max effort, default sampling, avg 5 trials)."
claude-opus-4.6,simplebench,67.6,https://lmcouncil.ai/benchmarks,{},,,dropped,"lmcouncil.ai scores are JS-rendered; could not extract value. R5g(c). DROPPED (R5h): random third-party blog/aggregator, no primary trail | DROPPED (R5g-c stale, never verified, source not audited in any pass)"
claude-sonnet-4.6,swe_bench_verified,79.6,https://www.anthropic.com/news/claude-sonnet-4-6,"{""effort"":""max"",""harness"":""simple bash+editor scaffold"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 10 trials)"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,Anthropic Sonnet 4.6 blog table: claude-sonnet-4.6 = 79.6.
claude-sonnet-4.6,osworld,72.5,https://www.anthropic.com/news/claude-sonnet-4-6,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,Anthropic Sonnet 4.6 blog table: claude-sonnet-4.6 = 72.5.
claude-sonnet-4.6,arc_agi_2,58.3,https://www.anthropic.com/news/claude-sonnet-4-6,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""120k thinking budget"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,Anthropic Sonnet 4.6 blog table: claude-sonnet-4.6 = 58.3.
claude-sonnet-4.6,gdpval_aa_elo,1633.0,https://www.anthropic.com/news/claude-sonnet-4-6,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,Anthropic Sonnet 4.6 blog table: claude-sonnet-4.6 = 1633.
claude-sonnet-4.6,aa_intelligence_index,51.0,https://artificialanalysis.ai/evaluations/artificial-analysis-intelligence-index,{},,,needs_review,"AA Intelligence Index: claude-sonnet-4-6-adaptive = 51.72 (rounds to 52), but BP has 51. 1-point discrepancy vs current AA data. Possible stale or truncation-vs-round difference. Current AA value: 51.72."
claude-sonnet-4.6,aime_2025,95.6,https://anthropic.com/claude-sonnet-4-6-system-card,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 10 trials)"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,"Claude Sonnet 4.6 System Card section 2.10: AIME 2025 95.6% without tools, avg over 10 trials, max effort. (Card notes potential contamination concern, see Section 2.2 of Opus 4.5 System Card.) [R5d: prior unverified value 97.0 from anthropic blog deleted.]"
claude-sonnet-4.6,browsecomp,74.7,https://www.anthropic.com/news/claude-sonnet-4-6,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic (web search + fetch + programmatic tool calling, context compaction at 50k up to 10M)""}",True,official_blog,verified,Anthropic Sonnet 4.6 blog table: claude-sonnet-4.6 = 74.7.
claude-sonnet-4.6,frontiermath,32.4,https://epoch.ai/benchmarks/frontiermath,"{""context"":""16k thinking"",""effort"":""n/a"",""mode"":""thinking"",""tools"":""none""}",False,leaderboard,verified,epoch.ai FrontierMath-2025-02-28-Private: claude-sonnet-4-6_16K = 32.4% (16k thinking). Canonical requests max effort but epoch.ai only has 16k thinking variant for Sonnet 4.6; no max entry exists. BP had 35.0.
claude-sonnet-4.6,gpqa_diamond,89.9,https://www.anthropic.com/news/claude-sonnet-4-6,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,Anthropic Sonnet 4.6 blog table: claude-sonnet-4.6 = 89.9.
claude-sonnet-4.6,hle,33.2,https://www.anthropic.com/news/claude-sonnet-4-6,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none (no-tools column)""}",True,official_blog,verified,Anthropic Sonnet 4.6 blog table: claude-sonnet-4.6 = 33.2.
claude-sonnet-4.6,humaneval,93.0,https://www.anthropic.com/news/claude-sonnet-4-6,{},,,needs_review,Anthropic blog benchmark table is a CDN PNG image; cannot extract text value. Web search shows conflicting values (90.8% vs JSON 93.0%). R5g(c).
claude-sonnet-4.6,ifeval,92.0,https://www.anthropic.com/news/claude-sonnet-4-6,{},,,needs_review,Anthropic blog benchmark table is a CDN PNG image; cannot extract text value. Web search shows 89.9% vs JSON 92.0%. R5g(c).
claude-sonnet-4.6,livecodebench,74.0,https://www.anthropic.com/news/claude-sonnet-4-6,{},,,needs_review,Anthropic blog benchmark table is a CDN PNG image; cannot extract text value. Web search shows 72.4% vs JSON 74.0%. R5g(c).
claude-sonnet-4.6,mmlu,90.0,https://www.anthropic.com/news/claude-sonnet-4-6,{},,,needs_review,Anthropic blog benchmark table is a CDN PNG image; cannot extract text value. Web search shows 89.7% vs JSON 90.0%. R5g(c).
claude-sonnet-4.6,mmlu_pro,87.0,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""maximum sequence length"",""effort"":""maximum reasoning effort"",""harness"":""Microsoft independent evaluation suite"",""judge"":""benchmark-specific Microsoft implementation"",""mode"":""thinking"",""notes"":""Sonnet 4.6 result generated by Microsoft's own evaluation."",""prompt_style"":""benchmark-specific"",""sampling"":""source does not state trial count"",""temperature"":""source does not state"",""tools"":""none; benchmark-provided tools where required""}",False,official_paper,verified,Displayed exactly as 87. Official Microsoft-reported result.
claude-sonnet-4.6,mmmu,74.2,https://automatio.ai/models/claude-sonnet-4-6,{},,,dropped,"Source is third-party blog (automatio.ai); R5h priority. DROPPED (R5h): random third-party blog/aggregator, no primary trail"
claude-sonnet-4.6,simpleqa,68.0,https://www.anthropic.com/news/claude-sonnet-4-6,{},,,needs_review,Anthropic blog benchmark table is a CDN PNG image; cannot extract text value. R5g(c).
claude-sonnet-4.6,tau_bench_retail,89.0,https://www.anthropic.com/news/claude-sonnet-4-6,{},,,dropped, | DROPPED: BP cell value 89.0 from anthropic.com/news/claude-sonnet-4-6 is mislabeled — Anthropic Sonnet 4.6 reports tau2-bench (v2) not tau-bench (v1). Use tau2_bench_retail (already verified at 91.7) instead.
claude-sonnet-4.6,tau_bench_telecom,97.0,https://www.anthropic.com/news/claude-sonnet-4-6,{},,,dropped, | DROPPED: BP cell value 97.0 from anthropic.com/news/claude-sonnet-4-6 is mislabeled — Anthropic Sonnet 4.6 reports tau2-bench (v2) not tau-bench (v1). Use tau2_bench_telecom (already verified at 97.9) instead.
claude-sonnet-4.6,arc_agi_1,86.5,https://anthropic.com/claude-sonnet-4-6-system-card,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""120k thinking tokens, ARC Prize Foundation private set"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,Claude Sonnet 4.6 System Card section 2.7: ARC-AGI-1 86.5% reported by ARC Prize Foundation with 120k thinking tokens and high effort on private set. matches_canonical=false (high effort vs canonical max). [R5d: prior unverified value 86.0 from arcprize.org leaderboard deleted.]
claude-sonnet-4.6,codeforces_rating,2010.0,https://www.nxcode.io/resources/news/claude-sonnet-4-6-complete-guide-benchmarks-pricing-2026,{},,,needs_review,Source is third-party blog (nxcode.io); R5h priority.
claude-sonnet-4.6,math_500,96.5,https://www.nxcode.io/resources/news/claude-sonnet-4-6-complete-guide-benchmarks-pricing-2026,{},,,needs_review,Source is third-party blog (nxcode.io); R5h priority.
claude-sonnet-4.6,mmmu_pro,74.5,https://www.anthropic.com/news/claude-sonnet-4-6,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""updated grading: separate Sonnet 4 grader; no 'think step-by-step' prefix"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,Anthropic Sonnet 4.6 blog table: claude-sonnet-4.6 = 74.5.
claude-sonnet-4.6,mrcr_v2,84.9,https://deepmind.google/models/gemini/pro/,"{""context"":""128k"",""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/pro/ Performance table (PROMOTED from candidate; prior unverified value 82 from https://awesomeagents.ai/leaderboards/long-context-benchmarks-leaderboard/ deleted per audit rule). DeepMind /models/gemini/pro/ cross-model Performance table: MRCR v2 8-needle 128k average 84.9%
claude-sonnet-4.6,swe_bench_pro,58.1,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""table8"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 8.1.A, page 115: Claude Sonnet 4.6; SWE-bench Pro [summary]=58.1. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=table8."
claude-sonnet-4.6,terminal_bench,59.1,https://www.anthropic.com/news/claude-sonnet-4-6,"{""effort"":""max"",""harness"":""Terminus-2 (1\u00d7 guaranteed/3\u00d7 ceiling)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5-15 trials)"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,Anthropic Sonnet 4.6 blog table: claude-sonnet-4.6 = 59.1.
claude-mythos,swe_bench_verified,93.9,https://www-cdn.anthropic.com/8b8380204f74670be75e81c820ca8dda846ab289.pdf,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic (bash/edit + thinking)""}",True,tech_report,verified,"Mythos sys card §6.4: 5 trials, standard config."
claude-mythos,swe_bench_pro,77.8,https://www-cdn.anthropic.com/8b8380204f74670be75e81c820ca8dda846ab289.pdf,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic""}",True,tech_report,verified,Mythos sys card §6.4: 5 trials.
claude-mythos,terminal_bench,82.0,https://www-cdn.anthropic.com/8b8380204f74670be75e81c820ca8dda846ab289.pdf,"{""effort"":""max"",""harness"":""Harbor scaffold + Terminus-2"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic""}",True,tech_report,verified,Mythos sys card §6.5.
claude-mythos,gpqa_diamond,94.5,https://www-cdn.anthropic.com/8b8380204f74670be75e81c820ca8dda846ab289.pdf,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Mythos sys card: 94.55%.
claude-mythos,hle,56.8,https://www-cdn.anthropic.com/8b8380204f74670be75e81c820ca8dda846ab289.pdf,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Mythos sys card Table 6.3.A: no-tools column. With-tools=64.7 not used.
claude-mythos,osworld,79.6,https://www-cdn.anthropic.com/8b8380204f74670be75e81c820ca8dda846ab289.pdf,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic""}",True,tech_report,verified,Mythos sys card Table 6.3.A.
claude-mythos,usamo_2025,97.6,https://www-cdn.anthropic.com/8b8380204f74670be75e81c820ca8dda846ab289.pdf,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 10 trials)"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,"Mythos sys card §6.8: max effort, no tools, avg 10 trials per problem. MathArena grading methodology (3 LLM judges incl. Gemini 3.1 Pro neutral rewriter)."
claude-opus-4.7,swe_bench_pro,64.3,https://www.anthropic.com/news/claude-opus-4-7,"{""effort"":""default"",""harness"":""simple bash+editor scaffold"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,Anthropic Opus 4.7 blog table: claude-opus-4.7 = 64.3%.
claude-opus-4.7,swe_bench_verified,87.6,https://www.anthropic.com/news/claude-opus-4-7,"{""effort"":""default"",""harness"":""simple bash+editor scaffold"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,Anthropic Opus 4.7 blog table: claude-opus-4.7 = 87.6%.
claude-opus-4.7,terminal_bench,69.4,https://www.anthropic.com/news/claude-opus-4-7,"{""effort"":""default"",""harness"":""Terminus-2 (1\u00d7 guaranteed/3\u00d7 ceiling, avg 5)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,Anthropic Opus 4.7 blog table: claude-opus-4.7 = 69.4%.
claude-opus-4.7,hle,46.9,https://www.anthropic.com/news/claude-opus-4-7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,Anthropic Opus 4.7 blog table: claude-opus-4.7 = 46.9%.
claude-opus-4.7,browsecomp,79.3,https://www.anthropic.com/news/claude-opus-4-7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,Anthropic Opus 4.7 blog table: claude-opus-4.7 = 79.3%.
claude-opus-4.7,gpqa_diamond,94.2,https://www.anthropic.com/news/claude-opus-4-7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,Anthropic Opus 4.7 blog table: claude-opus-4.7 = 94.2%.
claude-opus-4.7,arc_agi_1,92.0,https://arcprize.org/leaderboard,"{""effort"":""max"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""CoT"",""tools"":""none""}",False,leaderboard,verified,arcprize.org leaderboard audit: Claude 4.7 (Max) on leaderboard; canonical is default effort
claude-opus-4.7,arc_agi_2,75.8,https://arcprize.org/leaderboard,"{""effort"":""max"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""CoT"",""tools"":""none""}",False,leaderboard,verified,arcprize.org leaderboard audit: Claude 4.7 (Max) on leaderboard; canonical is default effort
claude-opus-4.7,frontiermath,43.8,https://openai.com/index/introducing-gpt-5-5/,{},False,third_party,verified,GPT-5.5 blog table: FrontierMath Tier 1–3=43.8%; third-party self-test by OpenAI
gemini-2.0-flash,mmlu_pro,77.6,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,"Gemini 2.0 Flash Model Card (April 15 2025): MMLU-Pro 77.6% [Promoted to verified per R5d, prior unverified value 76.4 from https://www.helicone.ai/blog/gemini-2.0-flash deleted.]"
gemini-2.0-flash,mmmu,71.7,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,"Gemini 2.0 Flash Model Card (April 15 2025): MMMU 71.7% [Promoted to verified per R5d, prior unverified value 70.7 from https://blog.google/technology/google-deepmind/gemini-model-updates-february-2025/ deleted.]"
gemini-2.0-flash,gpqa_diamond,60.1,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,"Gemini 2.0 Flash Model Card (April 15 2025): GPQA Diamond 60.1% [Promoted to verified per R5d, prior unverified value 74.2 from https://www.marktechpost.com/2025/01/21/google-ai-releases-gemini-2-0-flash-thinking-model/ deleted.]"
gemini-2.0-flash,aime_2024,73.3,https://www.marktechpost.com/2025/01/21/google-ai-releases-gemini-2-0-flash-thinking-model/,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gemini-2.0-flash,arc_agi_2,1.3,https://arcprize.org/leaderboard,"{""effort"":""n/a"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""non-thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""Base LLM"",""tools"":""none""}",True,leaderboard,verified,arcprize.org leaderboard audit: Gemini 2.0 Flash Base LLM on leaderboard
gemini-2.0-flash,arena_hard,50.0,https://github.com/lmarena/arena-hard-auto,"""Arena-Hard-v2.0 Creative Writing, Ensemble GPT-4.1+Gemini-2.5-Pro judge""",True,leaderboard,verified,
gemini-2.0-flash,ifeval,80.0,https://www.helicone.ai/blog/gemini-2.0-flash,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gemini-2.0-flash,mmlu,76.4,https://www.marktechpost.com/2025/01/21/google-ai-releases-gemini-2-0-flash-thinking-model-gemini-2-0-flash-thinking-exp-01-21-scoring-73-3-on-aime-math-and-74-2-on-gpqa-diamond-science-benchmarks/,{},,,dropped, | DROPPED: BP cell value 76.4 from marktechpost.com matches mmlu_pro=76.4 from same source — likely source mislabeled MMLU-Pro as MMLU. Standard MMLU 5-shot 14k not reported in Gemini 2.0 Flash model card. Use mmlu_pro cell instead.
gemini-2.0-flash,simpleqa,29.9,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,"Gemini 2.0 Flash Model Card (April 15 2025): SimpleQA (no search) 29.9% [Promoted to verified per R5d, prior unverified value 27.0 from https://www.helicone.ai/blog/gemini-2.0-flash deleted.]"
gemini-2.0-flash,bigcodebench,45.9,https://bigcode-bench.github.io/,"""BigCodeBench Complete Instruct pass@1""",True,leaderboard,verified,Source entry: Gemini-2.0-Flash-Exp (2025-02-05)
gemini-2.0-flash,humaneval,82.6,https://artificialanalysis.ai/models/gemini-2-0-flash,{},,,needs_review,"Reference URL is AA model page (gemini-2-0-flash). Current AA model page shows humaneval=0.904→90.4%, but BP has 82.6%. Major mismatch (7.8pp). Likely stale data; AA may have updated score. Current AA value: 90.4%."
gemini-2.0-flash,livecodebench,34.5,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,"Gemini 2.0 Flash Model Card (April 15 2025): LiveCodeBench v5 (UI 10/01/2024-02/01/2025) 34.5% [Promoted to verified per R5d, prior unverified value 45.2 from https://artificialanalysis.ai/models/gemini-2-0-flash deleted.]"
gemini-2.0-flash,math_500,83.9,https://artificialanalysis.ai/models/gemini-2-0-flash,{},,,needs_review,"Reference URL is AA model page (gemini-2-0-flash). Current AA model page shows math_500=0.93→93.0%, but BP has 83.9%. Major mismatch (9.1pp). Likely stale data; AA may have updated score. Current AA value: 93.0%."
gemini-2.0-flash,swe_bench_verified,42.0,https://artificialanalysis.ai/models/gemini-2-0-flash,{},,,needs_review,Reference URL is AA model page (gemini-2-0-flash). AA does not publish swe_bench_verified scores; field is not present in AA model data. Cannot verify 42 from AA source. Original source may be SWE-bench.com.
gemini-2.5-pro,aime_2025,88.0,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"DeepMind /models/gemini/flash/ Performance table: AIME 2025 no tools 88.0% [Promoted to verified, prior unverified value deleted per R5d.]"
gemini-2.5-pro,aime_2024,92.0,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: gemini-2.5-pro=92.0.
gemini-2.5-pro,gpqa_diamond,86.4,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"DeepMind /models/gemini/flash/ Performance table: GPQA Diamond 86.4% [Promoted to verified, prior unverified value deleted per R5d.]"
gemini-2.5-pro,hle,21.6,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"DeepMind /models/gemini/flash/ Performance table: HLE no tools 21.6% (Gemini 2.5 Pro Thinking) [Promoted to verified, prior unverified value deleted per R5d.]"
gemini-2.5-pro,swe_bench_verified,59.6,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"DeepMind /models/gemini/flash/ Performance table: SWE-bench Verified 59.6% [Promoted to verified, prior unverified value deleted per R5d.]"
gemini-2.5-pro,math_500,97.3,https://www.helicone.ai/blog/gemini-2.5-full-developer-guide,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gemini-2.5-pro,livecodebench,69.0,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-5-Pro-Model-Card.pdf,"{""effort"":""default"",""harness"":""AI Studio API"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (single attempt)"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,"Gemini 2.5 Pro Model Card (June 2025 GA): LiveCodeBench (UI 1/1/2025-5/1/2025) single attempt 69.0% [Promoted to verified per R5d, prior unverified value 70.4 from https://www.helicone.ai/blog/gemini-2.5-full-developer-guide deleted.]"
gemini-2.5-pro,simpleqa,54.0,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-5-Pro-Model-Card.pdf,"{""effort"":""default"",""harness"":""AI Studio API"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (single attempt)"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,"Gemini 2.5 Pro Model Card (June 2025 GA): SimpleQA single attempt 54.0% [Promoted to verified per R5d, prior unverified value 52.9 from https://www.helicone.ai/blog/gemini-2.5-full-developer-guide deleted.]"
gemini-2.5-pro,mmmu,82.0,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-5-Pro-Model-Card.pdf,"{""effort"":""default"",""harness"":""AI Studio API"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (single attempt)"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,"Gemini 2.5 Pro Model Card (June 2025 GA): MMMU single attempt pass@1 82.0% (GA; was 81.7 from Experimental 03-25 column of same card, mislabeled) [Promoted to verified per R5d, prior unverified value 81.7 from https://modelcards.withgoogle.com/assets/documents/gemini-2.5-pro.pdf deleted.]"
gemini-2.5-pro,mmlu,89.8,https://modelcards.withgoogle.com/assets/documents/gemini-2.5-pro.pdf,{},,,dropped," | DROPPED: BP cell value 89.8 from modelcards.withgoogle.com/.../gemini-2.5-pro.pdf is actually Global MMLU Lite from the Experimental 03-25 column, NOT standard MMLU 5-shot 14k. Source mislabel + wrong-version. Standard MMLU 5-shot is not reported in this card."
gemini-2.5-pro,arena_hard,90.8,https://github.com/lmarena/arena-hard-auto,"""Arena-Hard-v2.0 Creative Writing, Ensemble GPT-4.1+Gemini-2.5-Pro judge""",True,leaderboard,verified,
gemini-2.5-pro,chatbot_arena_elo,1448.0,https://lmarena.ai/leaderboard/text,{},True,leaderboard,verified,lmarena.ai text arena (fetched 2026-04-29). Arena model name: gemini-2.5-pro (rank 42). ELO updates continuously; score reflects latest available.
gemini-2.5-pro,aa_intelligence_index,60.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2 model card: gemini-2.5-pro=60.
gemini-2.5-pro,arc_agi_2,4.9,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"DeepMind /models/gemini/flash/ Performance table: ARC-AGI-2 ARC Prize Verified 4.9% [Promoted to verified, prior unverified value deleted per R5d.]"
gemini-2.5-pro,codeforces_rating,2001.0,https://huggingface.co/Qwen/Qwen3-235B-A22B-Thinking-2507,{},True,model_card,verified,
gemini-2.5-pro,frontiermath,14.1,https://epoch.ai/benchmarks/frontiermath,"{""effort"":""default"",""mode"":""thinking"",""tools"":""none""}",True,leaderboard,verified,epoch.ai FrontierMath-2025-02-28-Private: gemini-2.5-pro (Gemini 2.5 Pro Jun 2025) = 14.1%. Canonical model-id gemini-2.5-pro matches. BP had 5.0 (stale/incorrect).
gemini-2.5-pro,hmmt_feb_2025,82.5,https://arxiv.org/abs/2504.21318,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4-reasoning paper Table 1 (gemini-2.5-pro column): hmmt_2025=82.5 (re-sourced from https://arxiv.org/abs/2504.21318, matches BP value)"
gemini-2.5-pro,ifeval,91.5,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: gemini-2.5-pro=91.5.
gemini-2.5-pro,imo_2025,31.55,https://matharena.ai/?comp=imo--imo_2025,"""thinking:default""",True,leaderboard,verified,
gemini-2.5-pro,matharena_apex_2025,0.5,https://www.vellum.ai/blog/google-gemini-3-benchmarks,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gemini-2.5-pro,mmlu_pro,86.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2 model card: gemini-2.5-pro=86.
gemini-2.5-pro,usamo_2025,24.0,https://files.sri.inf.ethz.ch/matharena/usamo_report.pdf,{},True,academic_paper,verified,
gemini-2.5-pro,terminal_bench,32.6,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""Terminus-2"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"DeepMind /models/gemini/flash/ Performance table: Terminal-Bench 2.0 Terminus-2 32.6% [Promoted to verified, prior unverified value deleted per R5d.]"
gemini-2.5-pro,arc_agi_1,27.6,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: gemini-2.5-pro=27.6.
gemini-2.5-pro,brumo_2025,90.0,https://matharena.ai/?comp=brumo--brumo_2025,"""thinking:default""",True,leaderboard,verified,
gemini-2.5-pro,cmimc_2025,58.13,https://matharena.ai/?comp=cmimc--cmimc_2025,"""thinking:default""",True,leaderboard,verified,
gemini-2.5-pro,critpt,2.0,https://github.com/CritPt-Benchmark/CritPt,"""CritPt leaderboard, \u03b5=1.5, % tasks solved""",True,leaderboard,verified,
gemini-2.5-pro,hmmt_nov_2025,80.0,https://matharena.ai/?comp=hmmt--hmmt_nov_2025,"""thinking:default""",True,leaderboard,verified,
gemini-2.5-pro,ifbench,49.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2 model card: gemini-2.5-pro=49.
gemini-2.5-pro,livebench,57.5,https://livebench.ai/,"{""effort"":""default"",""mode"":""thinking"",""tools"":""n/a""}",True,leaderboard,verified,livebench.ai table_2026_01_08 (latest as of 2026-04-28): gemini-2.5-pro-06-05-highthinking avg=57.5. Canonical thinking+default matches. BP had 58.3 (rounding difference).
gemini-2.5-pro,mrcr_v2,58.0,https://deepmind.google/models/gemini/flash/,"{""context"":""128k"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"DeepMind /models/gemini/flash/ Performance table: MRCR v2 8-needle 128k 58.0% [Promoted to verified, prior unverified value deleted per R5d.]"
gemini-2.5-pro,simplebench,62.4,https://lmcouncil.ai/benchmarks,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail | DROPPED (R5g-c stale, never verified, source not audited in any pass)"
gemini-2.5-pro,smt_2025,84.91,https://matharena.ai/?comp=smt--smt_2025,"""thinking:default""",True,leaderboard,verified,
gemini-2.5-pro,terminal_bench_1,25.3,https://www.tbench.ai/leaderboard/terminal-bench/1.0,{},,leaderboard,verified,"Confirmed 25.3% on tbench.ai Terminal-Bench 1.0 leaderboard (Terminus 1 scaffold, 2025-05-15)."
gemini-2.5-flash,aime_2025,72.0,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"DeepMind /models/gemini/flash/ Performance table: AIME 2025 no tools 72.0% [Promoted to verified, prior unverified value deleted per R5d.]"
gemini-2.5-flash,arena_hard,83.9,https://github.com/lmarena/arena-hard-auto,"""Arena-Hard-v2.0 Creative Writing, Ensemble GPT-4.1+Gemini-2.5-Pro judge""",True,leaderboard,verified,
gemini-2.5-flash,aime_2024,88.0,https://modelcards.withgoogle.com/assets/documents/gemini-2.5-flash.pdf,{},,,dropped,
gemini-2.5-flash,livecodebench,63.9,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-5-Flash-Model-Card.pdf,"{""effort"":""default"",""harness"":""AI Studio API"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.5 Flash Model Card (Dec 2025) GA Thinking column: LiveCodeBench v5 single attempt 63.9
gemini-2.5-flash,codeforces_rating,1995.0,https://huggingface.co/Qwen/Qwen3-30B-A3B-Thinking-2507,{},True,model_card,verified,
gemini-2.5-flash,gpqa_diamond,82.8,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"DeepMind /models/gemini/flash/ Performance table: GPQA Diamond 82.8% [Promoted to verified, prior unverified value deleted per R5d.]"
gemini-2.5-flash,hmmt_feb_2025,64.2,https://huggingface.co/Qwen/Qwen3-30B-A3B-Thinking-2507,{},True,model_card,verified,
gemini-2.5-flash,ifeval,84.3,https://huggingface.co/Qwen/Qwen3-30B-A3B-Instruct-2507,{},True,model_card,verified,
gemini-2.5-flash,mmlu,85.0,https://deepmind.google/technologies/gemini/flash/,{},,,dropped,
gemini-2.5-flash,mmlu_pro,81.1,https://huggingface.co/Qwen/Qwen3-30B-A3B-Instruct-2507,{},True,model_card,verified,
gemini-2.5-flash,terminal_bench,16.9,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""Terminus-2"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"DeepMind /models/gemini/flash/ Performance table: Terminal-Bench 2.0 Terminus-2 16.9% [Promoted to verified, prior unverified value deleted per R5d.]"
gemini-2.5-flash,arc_agi_1,33.3,https://arcprize.org/leaderboard,"{""effort"":""default"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""CoT"",""tools"":""none""}",True,leaderboard,verified,arcprize.org leaderboard audit: Gemini 2.5 Flash (Preview) on leaderboard
gemini-2.5-flash,arc_agi_2,2.5,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"DeepMind /models/gemini/flash/ Performance table: ARC-AGI-2 ARC Prize Verified 2.5% [Promoted to verified, prior unverified value deleted per R5d.]"
gemini-2.5-flash,brumo_2025,83.33,https://matharena.ai/?comp=brumo--brumo_2025,"""thinking:default""",True,leaderboard,verified,
gemini-2.5-flash,cmimc_2025,51.88,https://matharena.ai/?comp=cmimc--cmimc_2025,"""thinking:default""",True,leaderboard,verified,
gemini-2.5-flash,critpt,1.1,https://github.com/CritPt-Benchmark/CritPt,"""CritPt leaderboard, \u03b5=1.5, % tasks solved""",True,leaderboard,verified,
gemini-2.5-flash,hle,11.0,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"DeepMind /models/gemini/flash/ Performance table: HLE no tools 11.0% (Gemini 2.5 Flash Thinking) [Promoted to verified, prior unverified value deleted per R5d.]"
gemini-2.5-flash,humaneval,90.2,https://llm-stats.com/models/gemini-2.5-flash,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gemini-2.5-flash,livebench,46.9,https://livebench.ai/,"{""effort"":""default"",""mode"":""thinking"",""tools"":""n/a""}",False,leaderboard,verified,livebench.ai table_2026_01_08 (latest as of 2026-04-28): gemini-2.5-flash-06-05-highthinking avg=46.9. Canonical requires Dec 2025 GA version; livebench has Jun 2025 preview (46.9) and Sep 2025 preview (52.3). Used Jun 2025 as closest to model release. BP had 47.7. See review.
gemini-2.5-flash,math_500,95.2,https://llm-stats.com/models/gemini-2.5-flash,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gemini-2.5-flash,mmmu,79.7,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-5-Flash-Model-Card.pdf,"{""effort"":""default"",""harness"":""AI Studio API"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.5 Flash Model Card (Dec 2025) GA Thinking column: MMMU single attempt 79.7. [R5d: prior unverified value 73.5 from https://llm-stats.com/models/gemini-2.5-flash deleted.]
gemini-2.5-flash,simpleqa,26.9,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-5-Flash-Model-Card.pdf,"{""effort"":""default"",""harness"":""AI Studio API"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.5 Flash Model Card (Dec 2025) GA Thinking column: SimpleQA 26.9. [R5d: prior unverified value 28.1 from https://llm-stats.com/models/gemini-2.5-flash deleted.]
gemini-2.5-flash,smt_2025,75.47,https://matharena.ai/?comp=smt--smt_2025,"""thinking:default""",True,leaderboard,verified,
gemini-2.5-flash,swe_bench_verified,60.4,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"DeepMind /models/gemini/flash/ Performance table: SWE-bench Verified 60.4% [Promoted to verified, prior unverified value deleted per R5d.]"
gemini-2.5-flash,terminal_bench_1,16.8,https://www.tbench.ai/leaderboard/terminal-bench/1.0,{},,leaderboard,verified,"Confirmed 16.8% on tbench.ai Terminal-Bench 1.0 leaderboard (Terminus 1 scaffold, 2025-05-17)."
gemma-3-27b,mmlu_pro,67.5,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report Table 6 IT 27B: MMLU-Pro 67.5
gemma-3-27b,gpqa_diamond,42.4,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report Table 6 IT 27B: GPQA Diamond 42.4
gemma-3-27b,livecodebench,29.7,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report Table 6 IT 27B: LiveCodeBench 29.7
gemma-3-27b,humaneval,87.8,https://www.emergentmind.com/topics/gemma-3-27b-it,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gemma-3-27b,chatbot_arena_elo,1338.0,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report Table 6 IT 27B: Chatbot Arena Elo 1338 (Table 5)
gemma-3-27b,mmlu,76.9,https://arxiv.org/abs/2503.19786,{},,,dropped, | DROPPED: BP value 76.9 with cell URL arxiv:2503.19786 — Table 6 IT does not include MMLU; pre-trained Table 9 shows 75.2. Value 76.9 not present in paper. Ghost cell per R5g(a).
gemma-3-27b,arena_hard,69.9,https://github.com/lmarena/arena-hard-auto,"""Arena-Hard-v2.0 Creative Writing, Ensemble GPT-4.1+Gemini-2.5-Pro judge""",True,leaderboard,verified,
gemma-3-27b,gsm8k,92.3,https://arxiv.org/abs/2503.19786,{},,,dropped, | DROPPED: BP value 92.3 with cell URL arxiv:2503.19786 — Gemma 3 paper Table 6 IT does not report GSM8K; pre-trained 27B GSM8K = 74.6. Value 92.3 not present. Ghost cell per R5g(a).
gemma-3-27b,aime_2024,22.0,https://llm-stats.com/benchmarks,{},,,dropped, | DROPPED: BP cell URL llm-stats with value 22; Gemma 3 paper does not report AIME 2024. Third-party ghost; canonical paper has no entry to verify against.
gemma-3-27b,ifeval,90.4,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-3-27b,math_500,78.0,https://llm-stats.com/benchmarks,{},,,dropped," | DROPPED: BP cell URL llm-stats with value 78; Gemma 3 paper Table 6 reports MATH (full set, score 89.0) not MATH-500 subset. Wrong-benchmark ghost."
gemma-3-27b,simpleqa,10.0,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report Table 6 IT 27B: SimpleQA 10.0 (was 22 from llm-stats). [R5d: prior unverified value 22 from https://llm-stats.com/benchmarks deleted.]
gemma-3-27b,swe_bench_verified,32.0,https://llm-stats.com/benchmarks,{},,,dropped, | DROPPED: BP cell URL llm-stats with value 32; Gemma 3 paper does not report SWE-bench Verified. Third-party ghost.
gemini-3-pro,gpqa_diamond,91.9,https://deepmind.google/models/gemini/pro/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/pro/ Performance table: GPQA Diamond no tools 91.9%
gemini-3-pro,aime_2025,95.0,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"DeepMind /models/gemini/flash/ Performance table: AIME 2025 no tools 95.0% [Promoted to verified, prior unverified value deleted per R5d.]"
gemini-3-pro,swe_bench_verified,76.2,https://deepmind.google/models/gemini/pro/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/pro/ Performance table: SWE-Bench Verified single attempt 76.2%
gemini-3-pro,hle,37.5,https://deepmind.google/models/gemini/pro/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/pro/ Performance table: HLE no tools 37.5%
gemini-3-pro,arc_agi_2,31.1,https://deepmind.google/models/gemini/pro/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/pro/ Performance table: ARC-AGI-2 ARC Prize Verified 31.1%
gemini-3-pro,mmmu_pro,81.0,https://deepmind.google/models/gemini/pro/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/pro/ Performance table: MMMU-Pro no tools 81.0%
gemini-3-pro,mmlu_pro,90.1,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: gemini-3-pro=90.1.
gemini-3-pro,mrcr_v2,77.0,https://deepmind.google/models/gemini/pro/,"{""context"":""128k"",""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/pro/ Performance table: MRCR v2 8-needle 128k average 77.0%
gemini-3-pro,simpleqa,72.1,https://officechai.com/ai/gemini-3-1-pro-benchmarks/,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gemini-3-pro,chatbot_arena_elo,1486.0,https://ernie.baidu.com/blog/posts/ernie-5.1-0508-release/creative-capability.png,"{""leaderboard"":""Arena AI Text"",""metric"":""Arena score"",""snapshot_date"":""2026-04-30"",""style_control"":""on""}",False,official_blog,verified,"April 30, 2026 Text Arena snapshot; style control on. Research observation obs-055. Displayed rank 8. Identical asset is published in four page placements. The same PNG bytes are published under two filenames and four English/Chinese page placements."
gemini-3-pro,livecodebench,90.7,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: gemini-3-pro=90.7.
gemini-3-pro,browsecomp,59.2,https://deepmind.google/models/gemini/pro/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""web+code""}",True,official_blog,verified,DeepMind /models/gemini/pro/ Performance table (PROMOTED over prior third-party): BrowseComp Search+Python+Browse 59.2% (was 85.9 from vellum.ai which mislabeled 3.1 Pro number; demoted to candidate)
gemini-3-pro,tau_bench_telecom,99.3,https://www.vellum.ai/blog/google-gemini-3-benchmarks,{},,,dropped," | DROPPED: BP cell value 99.3 from vellum.ai is actually τ²-bench-telecom for 3.1 Pro (not 3 Pro, not τ-bench). Source mislabel + wrong-model. Use tau2_bench_telecom for 3-pro."
gemini-3-pro,mmlu,91.8,https://www.vellum.ai/blog/google-gemini-3-benchmarks,{},,,dropped, | DROPPED: BP cell value 91.8 from vellum.ai is actually MMMLU not MMLU. Source mislabel. Use mmmlu cell instead.
gemini-3-pro,humaneval,93.0,https://www.vellum.ai/blog/google-gemini-3-benchmarks,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gemini-3-pro,frontiermath,38.0,https://x.com/EpochAIResearch/status/1991945942174761050,{},,,needs_review,
gemini-3-pro,codeforces_rating,2708.0,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: gemini-3-pro=2708.
gemini-3-pro,swe_bench_pro,43.3,https://deepmind.google/models/gemini/pro/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/pro/ Performance table: SWE-Bench Pro single attempt 43.3%
gemini-3-pro,mmmu,87.2,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,model_card,verified,"Qwen3.5-397B-A17B HF model card cross-model table (Qwen self-test of 3rd-party model, mc=false): mmmu 87.2 (3rd-party Qwen self-test). [R5d: prior unverified value 87.51 from https://www.vellum.ai/blog/google-gemini-3-benchmarks deleted.]"
gemini-3-pro,math_500,97.3,https://artificialanalysis.ai/evaluations/math-500,{},,,needs_review,"Reference URL is AA math-500 eval page, but gemini-3-1-pro-preview is NOT present on that page; AA shows math_500=null for this model. Possible ghost score (R5g-a). Cannot verify 97.3% from AA source."
gemini-3-pro,ifeval,93.5,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,model_card,verified,"Qwen3.5-397B-A17B HF model card cross-model table (Qwen self-test of 3rd-party model, mc=false): ifeval 93.5 (3rd-party Qwen self-test). [R5d: prior unverified value 90.0 from https://www.vellum.ai/blog/google-gemini-3-benchmarks deleted.]"
gemini-3-pro,matharena_apex_2025,24.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gemini-3-pro=24.5.
gemini-3-pro,osworld,55.0,https://www.vellum.ai/blog/google-gemini-3-benchmarks,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gemini-3-pro,scicode,56.0,https://deepmind.google/models/gemini/pro/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/pro/ Performance table: SciCode 56% (rounded; existing 56.1 within rounding)
gemini-3-pro,terminal_bench,56.9,https://deepmind.google/models/gemini/pro/,"{""effort"":""high"",""harness"":""Terminus-2"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/pro/ Performance table: Terminal-Bench 2.0 Terminus-2 56.9%
gemini-3-pro,aime_2024,97.0,https://www.vellum.ai/blog/google-gemini-3-benchmarks,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gemini-3-pro,arc_agi_1,85.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gemini-3-pro=85.0.
gemini-3-pro,brumo_2025,98.33,https://matharena.ai/?comp=brumo--brumo_2025,"""thinking:high""",True,leaderboard,verified,
gemini-3-pro,cmimc_2025,90.0,https://matharena.ai/?comp=cmimc--cmimc_2025,"""thinking:high""",True,leaderboard,verified,
gemini-3-pro,critpt,9.1,https://github.com/CritPt-Benchmark/CritPt,"""CritPt leaderboard, \u03b5=1.5, % tasks solved""",True,leaderboard,verified,
gemini-3-pro,hmmt_nov_2025,93.3,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: gemini-3-pro=93.3.
gemini-3-pro,livebench,73.5,https://livebench.ai/,"{""effort"":""high"",""mode"":""thinking"",""tools"":""n/a""}",True,leaderboard,verified,livebench.ai table_2026_01_08 (latest as of 2026-04-28): gemini-3-pro-preview-11-2025-high avg=73.5. Canonical thinking+high matches. BP had 73.4 (0.1 rounding diff).
gemini-3-pro,smt_2025,93.4,https://matharena.ai/?comp=smt--smt_2025,"""thinking:high""",True,leaderboard,verified,
gemini-3-pro,tau_bench_retail,88.5,https://artificialanalysis.ai/articles/gemini-3-pro-everything-you-need-to-know,{},,,needs_review,Reference URL is AA article page (gemini-3-pro-everything-you-need-to-know). AA only publishes aggregate tau2 scores; no retail-domain breakdown available in AA data. gemini-3-1-pro-preview tau2=0.956 aggregate. Cannot verify 88.5 from AA source.
gemini-3-flash,gpqa_diamond,90.4,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: GPQA Diamond no tools 90.4%
gemini-3-flash,mmlu,91.8,https://www.vellum.ai/blog/google-gemini-3-benchmarks,{},,,dropped, | DROPPED: BP cell 91.8 from vellum.ai is actually MMMLU (multilingual) not MMLU. Source mislabel. Use mmmlu cell instead.
gemini-3-flash,mmlu_pro,88.59,https://www.vellum.ai/blog/google-gemini-3-benchmarks,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gemini-3-flash,swe_bench_verified,78.0,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: SWE-bench Verified single attempt 78.0%
gemini-3-flash,hle,33.7,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: HLE no tools 33.7%
gemini-3-flash,mmmu_pro,81.2,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: MMMU-Pro 81.2%
gemini-3-flash,aime_2025,95.2,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: AIME 2025 no tools 95.2%
gemini-3-flash,mmmu,87.63,https://www.vellum.ai/blog/google-gemini-3-benchmarks,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gemini-3-flash,chatbot_arena_elo,1473.0,https://lmarena.ai/leaderboard/text,{},True,leaderboard,verified,lmarena.ai text arena (fetched 2026-04-29). Arena model name: gemini-3-flash (rank 15). ELO updates continuously; score reflects latest available.
gemini-3-flash,browsecomp,75.0,https://www.vellum.ai/blog/google-gemini-3-benchmarks,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gemini-3-flash,frontiermath,35.6,https://epoch.ai/benchmarks/frontiermath,"{""effort"":""default"",""mode"":""thinking"",""tools"":""none""}",True,leaderboard,verified,epoch.ai FrontierMath-2025-02-28-Private: gemini-3-flash-preview = 35.6% (Display: Gemini 3 Flash). BP had 30.0.
gemini-3-flash,livecodebench,90.8,https://artificialanalysis.ai/evaluations/livecodebench,"{""effort"":""default"",""harness"":""AA standard evaluation"",""mode"":""thinking"",""notes"":""AA slug: gemini-3-flash-reasoning; livecodebench=90.8%"",""sampling"":""pass@1""}",,third_party,verified,ArtificialAnalysis aggregator; used as fallback when no primary source available (R5d-aggregator-fallback). AA LiveCodeBench eval page: gemini-3-flash-reasoning livecodebench=0.908 → 90.8%. AA slug: gemini-3-flash-reasoning (reasoning/thinking variant).
gemini-3-flash,osworld,50.0,https://www.vellum.ai/blog/google-gemini-3-benchmarks,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gemini-3-flash,simpleqa,60.0,https://www.vellum.ai/blog/google-gemini-3-benchmarks,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gemini-3-flash,swe_bench_pro,34.63,https://scale.com/leaderboard/swe_bench_pro_public,{},False,leaderboard,verified, [Verified against Scale.com SWE-Bench Pro Public leaderboard (2026-04-29): gemini-3-flash = 34.63. Scale uses standardized evaluation scaffolding.]
gemini-3-flash,terminal_bench,47.6,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""Terminus-2"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"DeepMind /models/gemini/flash/ (PROMOTED over prior third-party): Terminal-Bench 2.0 Terminus-2 47.6% (was 51.7 from tbench leaderboard, demoted to candidate)"
gemini-3-flash,aime_2024,93.0,https://medium.com/@leucopsis/gemini-3-flash-preliminary-review-34e7420e3be7,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gemini-3-flash,arc_agi_1,84.7,https://arcprize.org/leaderboard,"{""effort"":""high"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""CoT"",""tools"":""none""}",False,leaderboard,verified,arcprize.org leaderboard audit: Gemini 3 Flash Preview (High) on leaderboard; canonical is default
gemini-3-flash,arc_agi_2,33.6,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: ARC-AGI-2 ARC Prize Verified 33.6%
gemini-3-flash,brumo_2025,100.0,https://matharena.ai/?comp=brumo--brumo_2025,"""thinking:default""",True,leaderboard,verified,
gemini-3-flash,cmimc_2025,90.62,https://matharena.ai/?comp=cmimc--cmimc_2025,"""thinking:default""",True,leaderboard,verified,
gemini-3-flash,codeforces_rating,2100.0,https://medium.com/@leucopsis/gemini-3-flash-preliminary-review-34e7420e3be7,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gemini-3-flash,hmmt_nov_2025,93.33,https://matharena.ai/?comp=hmmt--hmmt_nov_2025,"""thinking:default""",True,leaderboard,verified,
gemini-3-flash,ifeval,88.2,https://automatio.ai/models/gemini-3-flash,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gemini-3-flash,livebench,73.0,https://livebench.ai/,"{""effort"":""high"",""mode"":""thinking"",""tools"":""n/a""}",True,leaderboard,verified,livebench.ai table_2026_01_08 (latest as of 2026-04-28): gemini-3-flash-preview-high avg=73.0. Canonical thinking mode matches (high = default thinking for Gemini 3 Flash per canonical notes). BP had 72.4.
gemini-3-flash,matharena_apex_2025,15.62,https://matharena.ai/?comp=apex--apex_2025,"""thinking:default""",True,leaderboard,verified,
gemini-3-flash,smt_2025,92.92,https://matharena.ai/?comp=smt--smt_2025,"""thinking:default""",True,leaderboard,verified,
gemini-3-flash,tau_bench_retail,82.0,https://artificialanalysis.ai/articles/gemini-3-flash-everything-you-need-to-know,{},,,needs_review,Reference URL is AA article page (gemini-3-flash-everything-you-need-to-know). AA only publishes aggregate tau2 scores; no retail-domain breakdown available in AA data. gemini-3-flash tau2=N/A in AA (reasoning variant has other tau data). Cannot verify 82 from AA source.
gemini-3.1-pro,arc_agi_2,77.1,https://deepmind.google/models/gemini/pro/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/pro/ Performance table: ARC-AGI-2 ARC Prize Verified 77.1%
gemini-3.1-pro,gpqa_diamond,94.3,https://deepmind.google/models/gemini/pro/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/pro/ Performance table: GPQA Diamond no tools 94.3%
gemini-3.1-pro,hle,44.4,https://deepmind.google/models/gemini/pro/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/pro/ Performance table: HLE no tools 44.4%
gemini-3.1-pro,swe_bench_verified,80.6,https://deepmind.google/models/gemini/pro/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/pro/ Performance table: SWE-Bench Verified single attempt 80.6%
gemini-3.1-pro,scicode,59.0,https://deepmind.google/models/gemini/pro/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/pro/ Performance table: SciCode 59% (rounded; existing 58.9 within rounding)
gemini-3.1-pro,aime_2025,100.0,https://automatio.ai/models/gemini-3-1-pro,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gemini-3.1-pro,mmlu_pro,91.0,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): gemini-3.1-pro (per their own canonical config) = 91.0.
gemini-3.1-pro,mmmu_pro,80.5,https://deepmind.google/models/gemini/pro/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/pro/ Performance table: MMMU-Pro no tools 80.5%
gemini-3.1-pro,mathvision,89.8,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.6 model card: gemini-3.1-pro=89.8.
gemini-3.1-pro,mrcr_v2,84.9,https://deepmind.google/models/gemini/pro/,"{""context"":""128k"",""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/pro/ Performance table: MRCR v2 8-needle 128k average 84.9%
gemini-3.1-pro,simpleqa,72.1,https://officechai.com/ai/gemini-3-1-pro-benchmarks/,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gemini-3.1-pro,ifeval,96.1,https://longcat.chat/blog/longcat-2.0/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""Meituan unified in-house harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Measured in-house by Meituan unless otherwise noted."",""prompt_style"":""default"",""sampling"":""source does not state"",""temperature"":""0.7; top_k=-1; top_p=0.95"",""tools"":""none""}",False,official_blog,verified,LongCat-2.0 official tech blog: IFEval = 96.1. Measured in-house by Meituan under the reported unified harness.
gemini-3.1-pro,critpt,17.7,https://artificialanalysis.ai/evaluations/critpt,"{""effort"":""default"",""harness"":""AA standard evaluation"",""mode"":""thinking"",""notes"":""Per AA evaluations/critpt: gemini-3-1-pro-preview critpt=17.7%"",""sampling"":""pass@1""}",,third_party,verified,ArtificialAnalysis aggregator; used as fallback when no primary source available (R5d-aggregator-fallback). AA CritPT eval page: gemini-3-1-pro-preview critpt=0.177 → 17.7%. AA slug: gemini-3-1-pro-preview.
gemini-3.1-pro,aa_intelligence_index,57.0,https://artificialanalysis.ai/evaluations/artificial-analysis-intelligence-index,"{""effort"":""default"",""harness"":""AA Intelligence Index composite"",""mode"":""thinking"",""notes"":""AA slug: gemini-3-1-pro-preview; II=57.18 rounds to 57""}",,third_party,verified,ArtificialAnalysis aggregator; used as fallback when no primary source available (R5d-aggregator-fallback). AA Intelligence Index: gemini-3-1-pro-preview = 57.18 (displayed as 57). AA slug: gemini-3-1-pro-preview.
gemini-3.1-pro,mmlu,92.6,https://deepmind.google/models/model-cards/gemini-3-1-pro/,{},,,dropped, | DROPPED: BP cell 92.6 from /models/model-cards/gemini-3-1-pro/ is actually MMMLU not MMLU. Source mislabel. Use mmmlu cell instead.
gemini-3.1-pro,browsecomp,85.9,https://deepmind.google/models/gemini/pro/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""web+code""}",True,official_blog,verified,DeepMind /models/gemini/pro/ Performance table: BrowseComp Search+Python+Browse 85.9%
gemini-3.1-pro,frontiermath,36.9,https://epoch.ai/benchmarks/frontiermath,"{""effort"":""n/a"",""mode"":""thinking"",""tools"":""none""}",False,leaderboard,verified,epoch.ai FrontierMath-2025-02-28-Private: gemini-3.1-pro-preview (Gemini 3.1 Pro Preview) = 36.9%. Only one variant available; canonical effort=high not confirmed. BP had 40.0.
gemini-3.1-pro,chatbot_arena_elo,1493.0,https://lmarena.ai/leaderboard/text,{},True,leaderboard,verified,lmarena.ai text arena (fetched 2026-04-29). Arena model name: gemini-3.1-pro-preview (rank 5). ELO updates continuously; score reflects latest available.
gemini-3.1-pro,osworld,72.0,https://www.trendingtopics.eu/gemini-3-1-pro-leads-most-benchmarks/,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gemini-3.1-pro,arc_agi_1,98.0,https://arcprize.org/leaderboard,"{""effort"":""default"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""CoT"",""tools"":""none""}",True,leaderboard,verified,"arcprize.org leaderboard audit: Gemini 3.1 Pro (Preview, 2026-02-19) on leaderboard"
gemini-3.1-pro,codeforces_rating,3052.0,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): gemini-3.1-pro (per their own canonical config) = 3052.
gemini-3.1-pro,humaneval,95.0,https://www.digitalapplied.com/blog/google-gemini-3-1-pro-benchmarks-pricing-guide,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gemini-3.1-pro,livecodebench,91.7,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): gemini-3.1-pro (per their own canonical config) = 91.7.
gemini-3.1-pro,aime_2024,98.0,https://deepmind.google/models/model-cards/gemini-3-1-pro/,{},,,dropped,
gemini-3.1-pro,aime_2026,98.2,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5.1 blog: gemini-3.1-pro=98.2.
gemini-3.1-pro,math_500,98.5,https://www.digitalapplied.com/blog/google-gemini-3-1-pro-benchmarks-pricing-guide,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
gemini-3.1-pro,matharena_apex_2025,60.9,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): gemini-3.1-pro (per their own canonical config) = 60.9.
gemini-3.1-pro,mmmu,87.5,https://deepmind.google/models/model-cards/gemini-3-1-pro/,{},,,dropped,
gemini-3.1-pro,simplebench,79.6,https://lmcouncil.ai/benchmarks,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail | DROPPED (R5g-c stale, never verified, source not audited in any pass)"
gemini-3.1-pro,swe_bench_pro,54.2,https://deepmind.google/models/gemini/pro/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/pro/ Performance table: SWE-Bench Pro single attempt 54.2%
gemini-3.1-pro,tau_bench_retail,90.5,https://deepmind.google/models/model-cards/gemini-3-1-pro/,{},,,dropped, | DROPPED: BP cell 90.5 from /models/model-cards/gemini-3-1-pro/ is actually τ²-bench-retail not τ-bench-retail. Use tau2_bench_retail (90.8).
gemini-3.1-pro,terminal_bench,68.5,https://deepmind.google/models/gemini/pro/,"{""effort"":""high"",""harness"":""Terminus-2"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/pro/ Performance table: Terminal-Bench 2.0 Terminus-2 68.5%
llama-4-scout,mmlu_pro,74.3,https://huggingface.co/meta-llama/Llama-4-Scout-17B-16E-Instruct,"{""harness"":""official-meta"",""judge"":""rule-based"",""prompt_style"":""5-shot pre-train"",""sampling"":""pass@1"",""temperature"":""0.0""}",False,model_card,verified,Instruction-tuned 0-shot MMLU-Pro per HF/GitHub model card and Meta blog comparison table.
llama-4-scout,gpqa_diamond,57.2,https://huggingface.co/meta-llama/Llama-4-Scout-17B-16E-Instruct,"{""harness"":""official-meta"",""judge"":""rule-based"",""prompt_style"":""zeroshot"",""sampling"":""pass@1"",""temperature"":""0.0""}",True,model_card,verified,Instruction-tuned GPQA Diamond per HF/GitHub model card and Meta blog comparison table.
llama-4-scout,mmmu,69.4,https://huggingface.co/meta-llama/Llama-4-Scout-17B-16E-Instruct,"{""harness"":""official-meta"",""judge"":""rule-based"",""prompt_style"":""zeroshot"",""sampling"":""pass@1"",""temperature"":""0.0""}",True,model_card,verified,Instruction-tuned MMMU per HF/GitHub model card and Meta blog comparison table.
llama-4-scout,livecodebench,32.8,https://huggingface.co/meta-llama/Llama-4-Scout-17B-16E-Instruct,"{""harness"":""official-meta"",""judge"":""rule-based"",""prompt_style"":""zeroshot"",""sampling"":""pass@1"",""temperature"":""0.0""}",True,model_card,verified,Instruction-tuned LiveCodeBench (10/01/2024-02/01/2025) per HF/GitHub model card and Meta blog comparison table.
llama-4-scout,chatbot_arena_elo,1350.0,,{},False,,dropped,"Scout Arena Elo not in Meta blog text (only Maverick 1417 mentioned). Need LMSYS leaderboard source. | RESOLVED: dropped after audit — original ref was bogus blog URL, secondary blog, or unverifiable JS-rendered Meta product page."
llama-4-scout,gsm8k,90.6,,{},False,,dropped,Meta has not published Scout GSM8K. The 90.6 in BP equals MGSM 90.6 from instruction-tuned table — mislabel.
llama-4-scout,humaneval,76.0,,{},False,,dropped,"Not reported by Meta in blog/MC for Scout. Need third-party source or drop. | RESOLVED: dropped after audit — original ref was bogus blog URL, secondary blog, or unverifiable JS-rendered Meta product page."
llama-4-scout,ifeval,81.0,,{},False,,dropped,"Not reported by Meta in blog/MC for Scout. Need third-party source or drop. | RESOLVED: dropped after audit — original ref was bogus blog URL, secondary blog, or unverifiable JS-rendered Meta product page."
llama-4-scout,arc_agi_1,0.5,,{},False,third_party,verified,ARC-AGI-1 is third-party leaderboard.
llama-4-scout,arc_agi_2,0.0,,{},False,third_party,verified,ARC-AGI-2 is third-party leaderboard.
llama-4-scout,arena_hard,72.0,,{},False,,dropped,"Arena Hard not in Meta blog text or PNG comparison tables for Scout. Need LMSYS/source. | RESOLVED: dropped after audit — original ref was bogus blog URL, secondary blog, or unverifiable JS-rendered Meta product page."
llama-4-scout,math_500,83.0,,{},False,,dropped,"Not in Meta blog, HF model card, or GitHub MODEL_CARD. Need third-party source or drop. | RESOLVED: dropped after audit — original ref was bogus blog URL, secondary blog, or unverifiable JS-rendered Meta product page."
llama-4-scout,mmlu,79.6,https://huggingface.co/meta-llama/Llama-4-Scout-17B-16E-Instruct,"{""harness"":""official-meta"",""judge"":""rule-based"",""prompt_style"":""5-shot pre-train"",""sampling"":""pass@1"",""temperature"":""0.0""}",False,model_card,verified,Pre-train 5-shot MMLU per HF/GitHub model card. BP previously had 88.5 (likely confused with MMLU-Pro).
llama-4-maverick,mmlu_pro,80.5,https://huggingface.co/meta-llama/Llama-4-Maverick-17B-128E-Instruct,"{""harness"":""official-meta"",""judge"":""rule-based"",""prompt_style"":""zeroshot"",""sampling"":""pass@1"",""temperature"":""0.0""}",True,model_card,verified,Instruction-tuned 0-shot MMLU-Pro per HF/GitHub model card and Meta blog comparison table.
llama-4-maverick,gpqa_diamond,69.8,https://huggingface.co/meta-llama/Llama-4-Maverick-17B-128E-Instruct,"{""harness"":""official-meta"",""judge"":""rule-based"",""prompt_style"":""zeroshot"",""sampling"":""pass@1"",""temperature"":""0.0""}",True,model_card,verified,Instruction-tuned GPQA Diamond per HF/GitHub model card and Meta blog comparison table.
llama-4-maverick,math_500,90.0,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 90.0. Third-party self-test by Mistral (their internal eval pipeline).
llama-4-maverick,mmmu,73.4,https://huggingface.co/meta-llama/Llama-4-Maverick-17B-128E-Instruct,"{""harness"":""official-meta"",""judge"":""rule-based"",""prompt_style"":""zeroshot"",""sampling"":""pass@1"",""temperature"":""0.0""}",True,model_card,verified,Instruction-tuned MMMU per HF/GitHub model card and Meta blog comparison table.
llama-4-maverick,livecodebench,43.4,https://huggingface.co/meta-llama/Llama-4-Maverick-17B-128E-Instruct,"{""harness"":""official-meta"",""judge"":""rule-based"",""prompt_style"":""zeroshot"",""sampling"":""pass@1"",""temperature"":""0.0""}",True,model_card,verified,Instruction-tuned LiveCodeBench (10/01/2024-02/01/2025) per HF/GitHub model card and Meta blog comparison table.
llama-4-maverick,chatbot_arena_elo,1417.0,https://ai.meta.com/blog/llama-4-multimodal-intelligence/,"{""harness"":""official-meta"",""judge"":""rule-based"",""prompt_style"":""zeroshot"",""sampling"":""pass@1"",""temperature"":""0.0""}",True,official_blog,verified,Meta blog text: Maverick 'scores ELO of 1417 on LMArena'.
llama-4-maverick,arena_hard,91.8,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 91.8. Third-party self-test by Mistral (their internal eval pipeline).
llama-4-maverick,humaneval,85.4,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 85.4. Third-party self-test by Mistral (their internal eval pipeline).
llama-4-maverick,ifeval,88.9,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 88.9. Third-party self-test by Mistral (their internal eval pipeline).
llama-4-maverick,simpleqa,20.0,,{},False,,dropped,"SimpleQA not in Meta blog/MC for Maverick. Need third-party source or drop. | RESOLVED: dropped after audit — original ref was bogus blog URL, secondary blog, or unverifiable JS-rendered Meta product page."
llama-4-maverick,aime_2024,52.0,,{},False,,dropped,"AIME 2024 not in Meta blog/MC for Maverick. Need third-party source or drop. | RESOLVED: dropped after audit — original ref was bogus blog URL, secondary blog, or unverifiable JS-rendered Meta product page."
llama-4-maverick,arc_agi_1,4.4,,{},False,third_party,verified,ARC-AGI-1 third-party leaderboard.
llama-4-maverick,arc_agi_2,0.0,,{},False,third_party,verified,ARC-AGI-2 third-party leaderboard.
llama-4-maverick,bigcodebench,49.7,,{},False,third_party,verified,BigCodeBench third-party leaderboard.
llama-4-maverick,critpt,0.0,,{},False,third_party,verified,CritPt third-party benchmark.
llama-4-maverick,mmlu,85.5,https://huggingface.co/meta-llama/Llama-4-Maverick-17B-128E-Instruct,"{""harness"":""official-meta"",""judge"":""rule-based"",""prompt_style"":""5-shot pre-train"",""sampling"":""pass@1"",""temperature"":""0.0""}",False,model_card,verified,Pre-train 5-shot MMLU per HF/GitHub model card. BP previously had 88.6 (likely confused with MMLU-Pro).
llama-4-maverick,swe_bench_verified,46.5,,{},False,third_party,verified,SWE-bench Verified third-party leaderboard.
llama-4-maverick,terminal_bench_1,15.5,,{},False,third_party,verified,Terminal-Bench third-party leaderboard.
llama-4-behemoth,math_500,95.0,https://ai.meta.com/blog/llama-4-multimodal-intelligence/,"{""harness"":""official-meta"",""judge"":""rule-based"",""prompt_style"":""zeroshot"",""sampling"":""pass@1"",""temperature"":""0.0""}",True,official_blog,verified,Meta blog comparison table and blog text: 'on several STEM benchmarks such as MATH-500'.
llama-4-behemoth,mmlu_pro,82.2,https://ai.meta.com/blog/llama-4-multimodal-intelligence/,"{""harness"":""official-meta"",""judge"":""rule-based"",""prompt_style"":""zeroshot"",""sampling"":""pass@1"",""temperature"":""0.0""}",True,official_blog,verified,Meta blog comparison table (Behemoth vs Claude Sonnet 3.7 / Gemini 2.0 Pro / GPT-4.5) and blog text.
llama-4-behemoth,gpqa_diamond,73.7,https://ai.meta.com/blog/llama-4-multimodal-intelligence/,"{""harness"":""official-meta"",""judge"":""rule-based"",""prompt_style"":""zeroshot"",""sampling"":""pass@1"",""temperature"":""0.0""}",True,official_blog,verified,Meta blog comparison table and blog text: 'on several STEM benchmarks such as MATH-500 and GPQA Diamond'.
llama-4-behemoth,mmmu,76.1,https://ai.meta.com/blog/llama-4-multimodal-intelligence/,"{""harness"":""official-meta"",""judge"":""rule-based"",""prompt_style"":""zeroshot"",""sampling"":""pass@1"",""temperature"":""0.0""}",True,official_blog,verified,Meta blog comparison table (Behemoth vs Claude Sonnet 3.7 / Gemini 2.0 Pro / GPT-4.5).
llama-4-behemoth,livecodebench,49.4,https://ai.meta.com/blog/llama-4-multimodal-intelligence/,"{""harness"":""official-meta"",""judge"":""rule-based"",""prompt_style"":""zeroshot"",""sampling"":""pass@1"",""temperature"":""0.0""}",True,official_blog,verified,Meta blog comparison table (Behemoth vs Claude Sonnet 3.7 / Gemini 2.0 Pro / GPT-4.5).
llama-4-behemoth,humaneval,85.0,,{},False,,dropped,"Not in Meta blog text or comparison PNG. Need source or drop. | RESOLVED: dropped after audit — original ref was bogus blog URL, secondary blog, or unverifiable JS-rendered Meta product page."
llama-4-behemoth,ifeval,86.0,,{},False,,dropped,"Not in Meta blog text or comparison PNG. Need source or drop. | RESOLVED: dropped after audit — original ref was bogus blog URL, secondary blog, or unverifiable JS-rendered Meta product page."
llama-4-behemoth,aime_2024,72.0,,{},False,,dropped,"Not in Meta blog text or comparison PNG. Need source or drop. | RESOLVED: dropped after audit — original ref was bogus blog URL, secondary blog, or unverifiable JS-rendered Meta product page."
llama-4-behemoth,mmlu,90.2,,{},False,,dropped,"Meta has not published Behemoth MMLU in any source (blog text, blog PNG comparison table, no model card released). 90.2 origin unknown."
llama-4-behemoth,simpleqa,44.0,,{},False,,dropped,"Not in Meta blog text or comparison PNG. Need source or drop. | RESOLVED: dropped after audit — original ref was bogus blog URL, secondary blog, or unverifiable JS-rendered Meta product page."
llama-4-behemoth,swe_bench_verified,55.0,,{},False,third_party,verified,SWE-bench Verified third-party leaderboard.
muse-spark,mmmu_pro,80.4,https://ai.meta.com/blog/introducing-muse-spark-msl/,"{""effort"":""default"",""harness"":""internal (Meta MSL)"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 4 runs)"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"Meta MSL blog main table: MMMU Pro 80.4, Thinking mode. Avg 4 runs."
muse-spark,hle,42.8,https://ai.meta.com/blog/introducing-muse-spark-msl/,"{""effort"":""default"",""harness"":""internal (Meta MSL)"",""judge"":""gpt-o3-mini (LLM-as-judge)"",""mode"":""thinking"",""notes"":""No-tools variant. Full 2,500-question HLE dataset."",""prompt_style"":""default"",""sampling"":""pass@1 (single-pass)"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"Meta MSL blog: HLE (No Tools) 42.8, Thinking mode."
muse-spark,arc_agi_2,42.5,https://ai.meta.com/blog/introducing-muse-spark-msl/,"{""effort"":""default"",""harness"":""kaggle/arc-agi2"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""pass@2 to match competition setup; 120-prompt public set"",""prompt_style"":""default"",""sampling"":""pass@2"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,"Meta MSL blog: ARC-AGI-2 42.5, Thinking mode. 120 public-set prompts, pass@2. matches_canonical=false: canonical may be pass@1."
muse-spark,gpqa_diamond,89.5,https://ai.meta.com/blog/introducing-muse-spark-msl/,"{""effort"":""default"",""harness"":""internal (Meta MSL)"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg of 4 runs"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"Meta MSL blog: GPQA Diamond 89.5, Thinking mode. Averaged over 4 runs."
muse-spark,swe_bench_verified,77.4,https://ai.meta.com/blog/introducing-muse-spark-msl/,"{""effort"":""default"",""harness"":""internal (Meta MSL)"",""judge"":""rule-based (official docker)"",""mode"":""thinking"",""notes"":""Bug-fixed 5 test issues per methodology doc (astropy #7606, sphinx 8595/9711/8269/8475)."",""prompt_style"":""default"",""sampling"":""avg 15 attempts"",""temperature"":""default"",""tools"":""bash + file ops (view/edit)""}",True,official_blog,verified,"Meta MSL blog: SWE-bench Verified 77.4, Thinking mode. Single-attempt avg 15 trials; 5 test issues fixed."
muse-spark,swe_bench_pro,52.4,https://ai.meta.com/blog/introducing-muse-spark-msl/,"{""effort"":""default"",""harness"":""internal (Meta MSL)"",""judge"":""rule-based (official eval scripts)"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg 4 attempts"",""temperature"":""default"",""tools"":""bash + file ops""}",True,official_blog,verified,"Meta MSL blog: SWE-bench Pro 52.4, Thinking mode."
muse-spark,terminal_bench,59.0,https://ai.meta.com/blog/introducing-muse-spark-msl/,"{""effort"":""default"",""harness"":""official terminal-bench-2"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg 15 attempts"",""temperature"":""default"",""tools"":""bash-only""}",True,official_blog,verified,"Meta MSL blog: Terminal-Bench 2.0 59.0, Thinking mode."
muse-spark,gdpval_aa_elo,1444.0,https://ai.meta.com/blog/introducing-muse-spark-msl/,"{""effort"":""default"",""harness"":""Artificial Analysis GDPval-AA Leaderboard"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Results from Artificial Analysis GDPval-AA leaderboard per Meta methodology"",""prompt_style"":""default"",""sampling"":""pass@1 (single-pass)"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,Meta MSL blog: GDPval-AA Elo 1444. Sourced from Artificial Analysis leaderboard.
grok-3-beta,gpqa_diamond,80.2,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: grok-3-beta=80.2.
grok-3-beta,mmlu_pro,79.9,https://x.ai/news/grok-3,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,xAI Grok 3 blog (cross-model non-reasoning table): grok-3-beta=79.9.
grok-3-beta,livecodebench,57.0,https://x.ai/news/grok-3,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,xAI Grok 3 blog (cross-model non-reasoning table): grok-3-beta=57.0.
grok-3-beta,chatbot_arena_elo,1402.0,https://x.com/lmarena_ai/status/1891706264800936307,{},,leaderboard,verified,
grok-3-beta,math_500,99.2,https://artificialanalysis.ai/evaluations/math-500,{},,third_party,verified,ArtificialAnalysis aggregator; used as fallback when no primary source available (R5d-aggregator-fallback)
grok-3-beta,aime_2025,77.3,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: grok-3-beta=77.3.
grok-3-beta,aime_2024,52.2,https://x.ai/news/grok-3,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,xAI Grok 3 blog (cross-model non-reasoning table): grok-3-beta=52.2.
grok-3-beta,simpleqa,43.6,https://x.ai/news/grok-3,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,xAI Grok 3 blog (cross-model non-reasoning table): grok-3-beta=43.6.
grok-3-beta,arc_agi_1,31.9,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: grok-3-beta=31.9.
grok-3-beta,arc_agi_2,0.0,https://arcprize.org/arc-agi/2/,{},,leaderboard,verified,
grok-3-beta,hle,18.2,https://artificialanalysis.ai/models/grok-3,{},,third_party,verified,ArtificialAnalysis aggregator; used as fallback when no primary source available (R5d-aggregator-fallback)
grok-3-beta,humaneval,87.3,https://artificialanalysis.ai/models/grok-3,{},,third_party,verified,ArtificialAnalysis aggregator; used as fallback when no primary source available (R5d-aggregator-fallback)
grok-3-beta,ifeval,83.4,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: grok-3-beta=83.4.
grok-3-beta,mmlu,88.0,https://artificialanalysis.ai/models/grok-3,{},,third_party,verified,ArtificialAnalysis aggregator; used as fallback when no primary source available (R5d-aggregator-fallback)
grok-3-beta,swe_bench_verified,48.5,https://artificialanalysis.ai/models/grok-3,{},,third_party,verified,ArtificialAnalysis aggregator; used as fallback when no primary source available (R5d-aggregator-fallback)
grok-3-beta,terminal_bench_1,17.5,https://www.tbench.ai/leaderboard/terminal-bench/1.0,{},,leaderboard,verified,
grok-4,gpqa_diamond,87.5,https://x.ai/news/grok-4,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"xAI Grok 4 blog: grok-4=87.5 (no tools, no Heavy)."
grok-4,hle,25.4,https://x.ai/news/grok-4,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"xAI Grok 4 blog: grok-4=25.4 (no tools, no Heavy)."
grok-4,arc_agi_2,15.9,https://x.ai/news/grok-4,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"xAI Grok 4 blog: grok-4=15.9 (no tools, no Heavy)."
grok-4,livecodebench,79.0,https://x.ai/news/grok-4,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"xAI Grok 4 blog: grok-4=79 (no tools, no Heavy)."
grok-4,swe_bench_verified,73.5,https://datasciencedojo.com/blog/grok-4/,{},,,dropped,"Dropped: third-party random blog, unreliable source (R5h-drop-third-party-blog)"
grok-4,aime_2025,91.7,https://x.ai/news/grok-4,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"xAI Grok 4 blog: grok-4=91.7 (no tools, no Heavy)."
grok-4,aa_intelligence_index,73.0,https://x.com/ArtificialAnlys/status/1943166841150644622,{},,third_party,verified,ArtificialAnalysis aggregator; used as fallback when no primary source available (R5d-aggregator-fallback)
grok-4,humaneval,88.0,https://automatio.ai/models/grok-4,{},,,dropped,"Dropped: third-party random blog, unreliable source (R5h-drop-third-party-blog)"
grok-4,simpleqa,48.0,https://automatio.ai/models/grok-4,{},,,dropped,"Dropped: third-party random blog, unreliable source (R5h-drop-third-party-blog)"
grok-4,ifeval,89.2,https://automatio.ai/models/grok-4,{},,,dropped,"Dropped: third-party random blog, unreliable source (R5h-drop-third-party-blog)"
grok-4,mmmu,75.0,https://automatio.ai/models/grok-4,{},,,dropped,"Dropped: third-party random blog, unreliable source (R5h-drop-third-party-blog)"
grok-4,mmmu_pro,59.2,https://automatio.ai/models/grok-4,{},,,dropped,"Dropped: third-party random blog, unreliable source (R5h-drop-third-party-blog)"
grok-4,codeforces_rating,2708.0,https://x.ai/news/grok-4,{},,,dropped,
grok-4,chatbot_arena_elo,1465.0,https://lmarena.ai/,{},,leaderboard,verified,
grok-4,aime_2024,94.0,https://www.llmrumors.com/news/grok-4-the-breakthrough-ai-model-that-changes-everything,{},,,dropped,"Dropped: third-party random blog, unreliable source (R5h-drop-third-party-blog)"
grok-4,arc_agi_1,66.7,https://arcprize.org/leaderboard,{},,leaderboard,verified,
grok-4,frontiermath,13.0,https://epoch.ai/blog/grok-4-math,{},,leaderboard,verified,
grok-4,imo_2025,11.9,https://matharena.ai/?comp=imo--imo_2025,{},,leaderboard,verified,
grok-4,mmlu,94.0,https://forgecode.dev/blog/grok-4-initial-impression/,{},,,dropped,"Dropped: third-party random blog, unreliable source (R5h-drop-third-party-blog)"
grok-4,mmlu_pro,87.0,https://artificialanalysis.ai/models/grok-4,{},,third_party,verified,ArtificialAnalysis aggregator; used as fallback when no primary source available (R5d-aggregator-fallback)
grok-4,terminal_bench,23.1,https://www.tbench.ai/leaderboard/terminal-bench/2.0,{},,leaderboard,verified,
grok-4,aa_lcr,68.0,https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning,{},,third_party,verified,ArtificialAnalysis aggregator; used as fallback when no primary source available (R5d-aggregator-fallback)
grok-4,brumo_2025,95.0,https://matharena.ai/?comp=brumo--brumo_2025,{},,leaderboard,verified,
grok-4,cmimc_2025,83.75,https://matharena.ai/?comp=cmimc--cmimc_2025,{},,leaderboard,verified,
grok-4,hmmt_nov_2025,88.33,https://matharena.ai/?comp=hmmt--hmmt_nov_2025,{},,leaderboard,verified,
grok-4,math_500,98.0,https://aitoolapp.com/grok-4/benchmarks/,{},,,dropped,"Dropped: third-party random blog, unreliable source (R5h-drop-third-party-blog)"
grok-4,matharena_apex_2025,2.08,https://matharena.ai/?comp=apex--apex_2025,{},,leaderboard,verified,
grok-4,osworld,48.0,https://aitoolapp.com/grok-4/benchmarks/,{},,,dropped,"Dropped: third-party random blog, unreliable source (R5h-drop-third-party-blog)"
grok-4,smt_2025,85.85,https://matharena.ai/?comp=smt--smt_2025,{},,leaderboard,verified,
grok-4,swe_bench_pro,46.5,https://artificialanalysis.ai/models/grok-4,{},,third_party,verified,ArtificialAnalysis aggregator; used as fallback when no primary source available (R5d-aggregator-fallback)
grok-4,terminal_bench_1,39.0,https://www.tbench.ai/leaderboard/terminal-bench/1.0,{},,leaderboard,verified,
grok-4,usamo_2025,37.5,https://x.ai/news/grok-4,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"xAI Grok 4 blog: grok-4=37.5 (no tools, no Heavy)."
grok-4.1,chatbot_arena_elo,1483.0,https://x.ai/news/grok-4-1,{},,,needs_review,
grok-4.1,swe_bench_verified,50.6,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"DeepMind /models/gemini/flash/ Performance table: SWE-bench Verified 50.6% [Promoted to verified, prior unverified value deleted per R5d.]"
grok-4.1,aime_2025,91.9,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"DeepMind /models/gemini/flash/ Performance table: AIME 2025 no tools 91.9% [Promoted to verified, prior unverified value deleted per R5d.]"
grok-4.1,mmmu_pro,63.0,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"DeepMind /models/gemini/flash/ Performance table: MMMU-Pro 63.0% [Promoted to verified, prior unverified value deleted per R5d.]"
grok-4.1,video_mmmu,87.6,https://llm-stats.com/benchmarks/videommmu,{},,,dropped,"Dropped: third-party random blog, unreliable source (R5h-drop-third-party-blog)"
grok-4.1,gpqa_diamond,84.3,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"DeepMind /models/gemini/flash/ Performance table: GPQA Diamond 84.3% [Promoted to verified, prior unverified value deleted per R5d.]"
grok-4.1,hle,17.6,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"DeepMind /models/gemini/flash/ Performance table: HLE no tools 17.6% (Grok 4.1 Fast Reasoning) [Promoted to verified, prior unverified value deleted per R5d.]"
grok-4.1,mmlu_pro,85.3,https://www.sentisight.ai/gemini-3-vs-grok-4-1-vs-chatgpt-5-1/,{},,,dropped,"Dropped: third-party random blog, unreliable source (R5h-drop-third-party-blog)"
grok-4.1,arc_agi_2,42.0,https://www.analyticsvidhya.com/blog/2025/11/gemini-3-vs-grok-4-1-best-ai-of-2025/,{},,,dropped,"Dropped: third-party random blog, unreliable source (R5h-drop-third-party-blog)"
grok-4.1,brumo_2025,97.5,https://matharena.ai/?comp=brumo--brumo_2025,{},,leaderboard,verified,
grok-4.1,cmimc_2025,84.38,https://matharena.ai/?comp=cmimc--cmimc_2025,{},,leaderboard,verified,
grok-4.1,codeforces_rating,2650.0,https://www.glbgpt.com/hub/chatgpt-5-1-vs-grok-4-1-2025/,{},,,dropped,"Dropped: third-party random blog, unreliable source (R5h-drop-third-party-blog)"
grok-4.1,frontiermath,38.0,https://www.analyticsvidhya.com/blog/2025/11/gemini-3-vs-grok-4-1-best-ai-of-2025/,{},,,dropped,"Dropped: third-party random blog, unreliable source (R5h-drop-third-party-blog)"
grok-4.1,hmmt_nov_2025,93.33,https://matharena.ai/?comp=hmmt--hmmt_nov_2025,{},,leaderboard,verified,
grok-4.1,humaneval,95.0,https://www.analyticsvidhya.com/blog/2025/11/gemini-3-vs-grok-4-1-best-ai-of-2025/,{},,,dropped,"Dropped: third-party random blog, unreliable source (R5h-drop-third-party-blog)"
grok-4.1,ifeval,91.0,https://www.sentisight.ai/gemini-3-vs-grok-4-1-vs-chatgpt-5-1/,{},,,dropped,"Dropped: third-party random blog, unreliable source (R5h-drop-third-party-blog)"
grok-4.1,livebench,60.0,https://livebench.ai/,{},,leaderboard,verified,
grok-4.1,livecodebench,82.0,https://www.sentisight.ai/gemini-3-vs-grok-4-1-vs-chatgpt-5-1/,{},,,dropped,"Dropped: third-party random blog, unreliable source (R5h-drop-third-party-blog)"
grok-4.1,math_500,98.5,https://www.sentisight.ai/gemini-3-vs-grok-4-1-vs-chatgpt-5-1/,{},,,dropped,"Dropped: third-party random blog, unreliable source (R5h-drop-third-party-blog)"
grok-4.1,matharena_apex_2025,5.21,https://matharena.ai/?comp=apex--apex_2025,{},,leaderboard,verified,
grok-4.1,osworld,52.0,https://www.analyticsvidhya.com/blog/2025/11/gemini-3-vs-grok-4-1-best-ai-of-2025/,{},,,dropped,"Dropped: third-party random blog, unreliable source (R5h-drop-third-party-blog)"
grok-4.1,simpleqa,55.0,https://www.glbgpt.com/hub/chatgpt-5-1-vs-grok-4-1-2025/,{},,,dropped,"Dropped: third-party random blog, unreliable source (R5h-drop-third-party-blog)"
grok-4.1,smt_2025,84.6,https://matharena.ai/?comp=smt--smt_2025,{},,leaderboard,verified,
grok-4.20,arc_agi_1,89.5,https://arcprize.org/leaderboard,{},,leaderboard,verified,
grok-4.20,arc_agi_2,65.1,https://arcprize.org/leaderboard,{},,leaderboard,verified,
deepseek-r1,gpqa_diamond,71.5,https://huggingface.co/deepseek-ai/DeepSeek-R1-0528,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek R1-0528 model card: deepseek-r1=71.5.
deepseek-r1,math_500,97.3,https://arxiv.org/abs/2501.12948,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,"DeepSeek R1 paper Table 3 (page 9, R1 final): 97.3."
deepseek-r1,aime_2024,79.8,https://huggingface.co/deepseek-ai/DeepSeek-R1-0528,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek R1-0528 model card: deepseek-r1=79.8.
deepseek-r1,mmlu,90.8,https://arxiv.org/abs/2501.12948,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,"DeepSeek R1 paper Table 3 (page 9, R1 final): 90.8."
deepseek-r1,mmlu_pro,84.0,https://huggingface.co/deepseek-ai/DeepSeek-R1-0528,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek R1-0528 model card: deepseek-r1=84.0.
deepseek-r1,swe_bench_verified,49.2,https://huggingface.co/deepseek-ai/DeepSeek-R1-0528,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,model_card,verified,DeepSeek R1-0528 model card: deepseek-r1=49.2.
deepseek-r1,livecodebench,65.9,https://arxiv.org/abs/2501.12948,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,"DeepSeek R1 paper Table 3 (page 9, R1 final): 65.9."
deepseek-r1,codeforces_rating,2029.0,https://arxiv.org/abs/2501.12948,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,"DeepSeek R1 paper Table 3 (page 9, R1 final): 2029."
deepseek-r1,arena_hard,92.3,https://arxiv.org/abs/2501.12948,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,"DeepSeek R1 paper Table 3 (page 9, R1 final): 92.3."
deepseek-r1,ifeval,83.3,https://arxiv.org/abs/2501.12948,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,"DeepSeek R1 paper Table 3 (page 9, R1 final): 83.3."
deepseek-r1,simpleqa,30.1,https://huggingface.co/deepseek-ai/DeepSeek-R1-0528,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek R1-0528 model card: deepseek-r1=30.1.
deepseek-r1,humaneval,96.1,https://arxiv.org/html/2501.12948v1,{},,,dropped, | DROPPED: cell URL is arxiv html v1 of R1 paper but R1 paper Table 8 (main) and Table 15 (distill) do not report HumanEval. Ghost cell per R5g(a).
deepseek-r1,aime_2025,70.0,https://huggingface.co/deepseek-ai/DeepSeek-R1-0528,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek R1-0528 model card: deepseek-r1=70.0.
deepseek-r1,hle,8.5,https://huggingface.co/deepseek-ai/DeepSeek-R1-0528,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek R1-0528 model card: deepseek-r1=8.5.
deepseek-r1,hmmt_feb_2025,41.7,https://huggingface.co/deepseek-ai/DeepSeek-R1-0528,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek R1-0528 model card: deepseek-r1=41.7.
deepseek-r1,chatbot_arena_elo,1398.0,https://lmarena.ai/leaderboard/text,{},True,leaderboard,verified,lmarena.ai text arena (fetched 2026-04-29). Arena model name: deepseek-r1 (rank 108). ELO updates continuously; score reflects latest available.
deepseek-r1,arc_agi_1,18.3,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: deepseek-r1=18.3.
deepseek-r1,gsm8k,97.3,https://arxiv.org/html/2501.12948v1,{},,,dropped, | DROPPED: cell URL is arxiv html v1 of R1 paper but R1 paper Table 8 reports MATH-500 not GSM8K. Ghost cell per R5g(a).
deepseek-r1,arc_agi_2,1.3,https://arcprize.org/leaderboard,"{""effort"":""default"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""CoT"",""tools"":""none""}",True,leaderboard,verified,arcprize.org leaderboard audit: Deepseek R1 on leaderboard
deepseek-r1,brumo_2025,80.83,https://matharena.ai/?comp=brumo--brumo_2025,{},,,verified, | Verified: matharena.ai shows 80.83%
deepseek-r1,critpt,1.1,https://github.com/CritPt-Benchmark/CritPt,"""CritPt leaderboard, \u03b5=1.5, % tasks solved""",True,leaderboard,verified,
deepseek-r1,ifbench,38.0,https://github.com/allenai/IFBench,"""IFBench score, 0-shot""",True,leaderboard,verified,
deepseek-r1,smt_2025,66.51,https://matharena.ai/?comp=smt--smt_2025,{},,,flagged," | MISMATCH: BP=66.51, current matharena=66.98; added candidate"
deepseek-r1,terminal_bench_1,5.7,https://www.tbench.ai/leaderboard/terminal-bench/1.0,{},,leaderboard,verified,"Confirmed 5.7% on tbench.ai Terminal-Bench 1.0 leaderboard (Terminus 1 scaffold, 2025-05-15)."
deepseek-r1,usamo_2025,4.76,https://matharena.ai/?comp=usamo--usamo_2025,{},,,verified, | Verified: matharena.ai shows 4.76%
deepseek-v3,mmlu,88.5,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v3=88.5.
deepseek-v3,gpqa_diamond,59.1,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v3=59.1.
deepseek-v3,math_500,90.2,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v3=90.2.
deepseek-v3,swe_bench_verified,42.0,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v3=42.0.
deepseek-v3,ifeval,86.1,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v3=86.1.
deepseek-v3,arena_hard,85.5,https://github.com/deepseek-ai/DeepSeek-V3,"""Arena-Hard open-ended generation, README Table""",True,tech_report,verified,
deepseek-v3,humaneval,82.6,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v3=82.6.
deepseek-v3,codeforces_rating,1134.0,https://arxiv.org/html/2501.12948v1,{},True,academic_paper,verified,
deepseek-v3,chatbot_arena_elo,1358.0,https://lmarena.ai/leaderboard/text,{},True,leaderboard,verified,lmarena.ai text arena (fetched 2026-04-29). Arena model name: deepseek-v3 (rank 148). ELO updates continuously; score reflects latest available.
deepseek-v3,gsm8k,89.3,https://github.com/deepseek-ai/DeepSeek-V3,"""GSM8K, 8-shot EM, README Table""",True,tech_report,verified,
deepseek-v3,bigcodebench,50.0,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper: bigcodebench=50.0 (matches BP, re-sourced)"
deepseek-v3-0324,gpqa_diamond,68.4,https://huggingface.co/deepseek-ai/DeepSeek-V3-0324,{},,,verified, | Verified: HF V3-0324 model card shows 68.4
deepseek-v3-0324,aime_2024,59.4,https://huggingface.co/deepseek-ai/DeepSeek-V3-0324,{},,,verified, | Verified: HF V3-0324 model card shows 59.4
deepseek-v3-0324,livecodebench,49.2,https://huggingface.co/deepseek-ai/DeepSeek-V3-0324,{},,,verified, | Verified: HF V3-0324 model card shows 49.2
deepseek-v3-0324,mmlu_pro,81.2,https://huggingface.co/deepseek-ai/DeepSeek-V3-0324,{},,,verified, | Verified: HF V3-0324 model card shows 81.2
deepseek-v3-0324,aime_2025,46.7,https://arxiv.org/abs/2507.20534,{},,,verified, | Verified: Kimi K2 paper (arXiv 2507.20534) Table 3 shows 46.7
deepseek-v3-0324,arena_hard,39.9,https://arxiv.org/abs/2507.20534,{},,,verified, | Verified: Kimi K2 paper (arXiv 2507.20534) Table 3 shows 39.9
deepseek-v3-0324,hle,5.2,https://arxiv.org/abs/2507.20534,{},,,verified, | Verified: Kimi K2 paper (arXiv 2507.20534) Table 3 shows 5.2
deepseek-v3-0324,hmmt_feb_2025,27.5,https://arxiv.org/abs/2507.20534,{},,,verified, | Verified: Kimi K2 paper (arXiv 2507.20534) Table 3 shows 27.5
deepseek-v3-0324,ifeval,81.1,https://arxiv.org/abs/2507.20534,{},,,verified, | Verified: Kimi K2 paper (arXiv 2507.20534) Table 3 shows 81.1
deepseek-v3-0324,math_500,94.0,https://arxiv.org/abs/2507.20534,{},,,verified, | Verified: Kimi K2 paper (arXiv 2507.20534) Table 3 shows 94.0
deepseek-v3-0324,mmlu,89.4,https://arxiv.org/abs/2507.20534,{},,,verified, | Verified: Kimi K2 paper (arXiv 2507.20534) Table 3 shows 89.4
deepseek-v3-0324,simpleqa,27.7,https://arxiv.org/abs/2507.20534,{},,,verified, | Verified: Kimi K2 paper (arXiv 2507.20534) Table 3 shows 27.7
deepseek-v3-0324,swe_bench_verified,38.8,https://arxiv.org/abs/2507.20534,{},,,verified, | Verified: Kimi K2 paper (arXiv 2507.20534) Table 3 shows 38.8
deepseek-v3-0324,codeforces_rating,1650.0,https://textcortex.com/post/deepseek-v3-review,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
deepseek-v3-0324,humaneval,85.0,https://textcortex.com/post/deepseek-v3-review,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
deepseek-r1-0528,aime_2025,87.5,https://huggingface.co/deepseek-ai/DeepSeek-R1-0528,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek R1-0528 model card: deepseek-r1-0528=87.5.
deepseek-r1-0528,aime_2024,91.4,https://huggingface.co/deepseek-ai/DeepSeek-R1-0528,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek R1-0528 model card: deepseek-r1-0528=91.4.
deepseek-r1-0528,gpqa_diamond,81.0,https://huggingface.co/deepseek-ai/DeepSeek-R1-0528,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek R1-0528 model card: deepseek-r1-0528=81.0.
deepseek-r1-0528,math_500,97.3,https://huggingface.co/deepseek-ai/DeepSeek-R1-0528,{},,,flagged, | FLAGGED: HF R1-0528 model card does not contain MATH-500; 97.3 is from DeepSeek-R1 (original) model card
deepseek-r1-0528,livecodebench,73.3,https://huggingface.co/deepseek-ai/DeepSeek-R1-0528,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek R1-0528 model card: deepseek-r1-0528=73.3.
deepseek-r1-0528,swe_bench_verified,57.6,https://huggingface.co/deepseek-ai/DeepSeek-R1-0528,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,model_card,verified,DeepSeek R1-0528 model card: deepseek-r1-0528=57.6.
deepseek-r1-0528,mmlu_pro,85.0,https://huggingface.co/deepseek-ai/DeepSeek-R1-0528,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek R1-0528 model card: deepseek-r1-0528=85.0.
deepseek-r1-0528,simpleqa,27.8,https://huggingface.co/deepseek-ai/DeepSeek-R1-0528,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek R1-0528 model card: deepseek-r1-0528=27.8.
deepseek-r1-0528,hle,17.7,https://huggingface.co/deepseek-ai/DeepSeek-R1-0528,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek R1-0528 model card: deepseek-r1-0528=17.7.
deepseek-r1-0528,codeforces_rating,1930.0,https://huggingface.co/deepseek-ai/DeepSeek-R1-0528,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek R1-0528 model card: deepseek-r1-0528=1930.
deepseek-r1-0528,hmmt_feb_2025,79.4,https://huggingface.co/deepseek-ai/DeepSeek-R1-0528,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek R1-0528 model card: deepseek-r1-0528=79.4.
deepseek-r1-0528,chatbot_arena_elo,1422.0,https://lmarena.ai/leaderboard/text,{},True,leaderboard,verified,lmarena.ai text arena (fetched 2026-04-29). Arena model name: deepseek-r1-0528 (rank 74). ELO updates continuously; score reflects latest available.
deepseek-r1-0528,frontiermath,,https://epoch.ai/benchmarks/frontiermath,{},,,dropped,"R5g-a ghost cell: DeepSeek-R1-0528 not found in epoch.ai FrontierMath-2025-02-28-Private leaderboard. Present in other epoch tasks (Tier 4, GPQA, MATH) but absent from main FM leaderboard. Original value 10.0 was unverified and source URL cannot substantiate it."
deepseek-r1-0528,ifeval,79.1,https://huggingface.co/Qwen/Qwen3-235B-A22B-Thinking-2507,{},True,model_card,verified,
deepseek-r1-0528,mmlu,90.8,https://medium.com/@leucopsis/deepseeks-new-r1-0528-performance-analysis-and-benchmark-comparisons-6440eac858d6,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
deepseek-r1-0528,arc_agi_1,21.2,https://arcprize.org/leaderboard,"{""effort"":""default"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""CoT"",""tools"":""none""}",True,leaderboard,verified,arcprize.org leaderboard audit: Deepseek R1 (05/28) on leaderboard
deepseek-r1-0528,arc_agi_2,1.1,https://arcprize.org/leaderboard,"{""effort"":""default"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""CoT"",""tools"":""none""}",True,leaderboard,verified,arcprize.org leaderboard audit: Deepseek R1 (05/28) on leaderboard
deepseek-r1-0528,brumo_2025,92.5,https://matharena.ai/?comp=brumo--brumo_2025,{},,,verified, | Verified: matharena.ai shows 92.5%
deepseek-r1-0528,cmimc_2025,69.38,https://matharena.ai/?comp=cmimc--cmimc_2025,{},,,verified, | Verified: matharena.ai shows 69.38%
deepseek-r1-0528,humaneval,85.6,https://medium.com/@leucopsis/deepseeks-new-r1-0528-performance-analysis-and-benchmark-comparisons-6440eac858d6,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
deepseek-r1-0528,imo_2025,6.85,https://matharena.ai/?comp=imo--imo_2025,{},,,verified, | Verified: matharena.ai shows 6.85%
deepseek-r1-0528,matharena_apex_2025,1.04,https://matharena.ai/?comp=apex--apex_2025,{},,,verified, | Verified: matharena.ai shows 1.04%
deepseek-r1-0528,smt_2025,83.02,https://matharena.ai/?comp=smt--smt_2025,{},,,verified, | Verified: matharena.ai shows 83.02%
deepseek-r1-0528,usamo_2025,30.06,https://matharena.ai/?comp=usamo--usamo_2025,{},,,verified, | Verified: matharena.ai shows 30.06%
deepseek-r1-distill-qwen-32b,aime_2024,72.6,https://arxiv.org/abs/2501.12948,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek R1 paper Table 15: deepseek-r1-distill-qwen-32b=72.6.
deepseek-r1-distill-qwen-32b,math_500,94.3,https://arxiv.org/abs/2501.12948,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek R1 paper Table 15: deepseek-r1-distill-qwen-32b=94.3.
deepseek-r1-distill-qwen-32b,gpqa_diamond,62.1,https://arxiv.org/abs/2501.12948,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek R1 paper Table 15: deepseek-r1-distill-qwen-32b=62.1.
deepseek-r1-distill-qwen-32b,livecodebench,57.2,https://arxiv.org/abs/2501.12948,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek R1 paper Table 15: deepseek-r1-distill-qwen-32b=57.2.
deepseek-r1-distill-qwen-32b,codeforces_rating,1691.0,https://arxiv.org/abs/2501.12948,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek R1 paper Table 15: deepseek-r1-distill-qwen-32b=1691.
deepseek-r1-distill-qwen-32b,aime_2025,72.6,https://arxiv.org/html/2501.12948v1,{},,,dropped," | DROPPED: cell URL arxiv html v1 of R1 paper, but R1 paper does NOT report distill-qwen-32b AIME 2025 (Table 13 only reports R1 main). Value 72.6 matches AIME 2024 score → likely scrape error. Ghost cell per R5g(a)."
deepseek-r1-distill-qwen-32b,humaneval,82.0,https://arxiv.org/html/2501.12948v1,{},,,dropped," | DROPPED: cell URL arxiv html v1, but R1 paper does not report distill HumanEval. Ghost cell."
deepseek-r1-distill-qwen-14b,aime_2024,69.7,https://arxiv.org/abs/2501.12948,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek R1 paper Table 15: deepseek-r1-distill-qwen-14b=69.7.
deepseek-r1-distill-qwen-14b,math_500,93.9,https://arxiv.org/abs/2501.12948,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek R1 paper Table 15: deepseek-r1-distill-qwen-14b=93.9.
deepseek-r1-distill-qwen-14b,codeforces_rating,1481.0,https://arxiv.org/abs/2501.12948,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek R1 paper Table 15: deepseek-r1-distill-qwen-14b=1481.
deepseek-r1-distill-qwen-14b,gpqa_diamond,59.1,https://arxiv.org/abs/2501.12948,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek R1 paper Table 15: deepseek-r1-distill-qwen-14b=59.1.
deepseek-r1-distill-qwen-14b,livecodebench,53.1,https://arxiv.org/abs/2501.12948,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek R1 paper Table 15: deepseek-r1-distill-qwen-14b=53.1.
deepseek-r1-distill-qwen-7b,aime_2024,55.5,https://arxiv.org/abs/2501.12948,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek R1 paper Table 15: deepseek-r1-distill-qwen-7b=55.5.
deepseek-r1-distill-qwen-7b,math_500,92.8,https://arxiv.org/abs/2501.12948,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek R1 paper Table 15: deepseek-r1-distill-qwen-7b=92.8.
deepseek-r1-distill-qwen-7b,gpqa_diamond,49.1,https://arxiv.org/abs/2501.12948,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek R1 paper Table 15: deepseek-r1-distill-qwen-7b=49.1.
deepseek-r1-distill-qwen-7b,livecodebench,37.6,https://arxiv.org/abs/2501.12948,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek R1 paper Table 15: deepseek-r1-distill-qwen-7b=37.6.
deepseek-r1-distill-qwen-7b,codeforces_rating,1189.0,https://arxiv.org/abs/2501.12948,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek R1 paper Table 15: deepseek-r1-distill-qwen-7b=1189.
deepseek-r1-distill-qwen-1.5b,aime_2024,28.9,https://arxiv.org/abs/2501.12948,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek R1 paper Table 15: deepseek-r1-distill-qwen-1.5b=28.9.
deepseek-r1-distill-qwen-1.5b,math_500,83.9,https://arxiv.org/abs/2501.12948,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek R1 paper Table 15: deepseek-r1-distill-qwen-1.5b=83.9.
deepseek-r1-distill-qwen-1.5b,gpqa_diamond,33.8,https://arxiv.org/abs/2501.12948,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek R1 paper Table 15: deepseek-r1-distill-qwen-1.5b=33.8.
deepseek-r1-distill-qwen-1.5b,livecodebench,13.2,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Qwen3 tech report (arxiv:2505.09388) Table 19 baseline column for deepseek-r1-distill-qwen-1.5b: LiveCodeBench v5 13.2. Third-party Qwen self-test, matches_canonical=false."
deepseek-r1-distill-qwen-1.5b,codeforces_rating,954.0,https://arxiv.org/abs/2501.12948,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek R1 paper Table 15: deepseek-r1-distill-qwen-1.5b=954.
deepseek-r1-distill-qwen-1.5b,aime_2025,22.8,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Qwen3 tech report (arxiv:2505.09388) Table 19 baseline column for deepseek-r1-distill-qwen-1.5b: AIME 2025 22.8. Third-party Qwen self-test, matches_canonical=false."
deepseek-r1-distill-qwen-1.5b,ifeval,39.9,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Qwen3 tech report (arxiv:2505.09388) Table 19 baseline column for deepseek-r1-distill-qwen-1.5b: IFEval strict 39.9. Third-party Qwen self-test, matches_canonical=false."
deepseek-r1-distill-qwen-1.5b,arena_hard,4.5,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Qwen3 tech report (arxiv:2505.09388) Table 19 baseline column for deepseek-r1-distill-qwen-1.5b: Arena-Hard 4.5. Third-party Qwen self-test, matches_canonical=false."
deepseek-r1-distill-llama-8b,aime_2024,50.4,https://arxiv.org/abs/2501.12948,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek R1 paper Table 15: deepseek-r1-distill-llama-8b=50.4.
deepseek-r1-distill-llama-8b,math_500,89.1,https://arxiv.org/abs/2501.12948,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek R1 paper Table 15: deepseek-r1-distill-llama-8b=89.1.
deepseek-r1-distill-llama-8b,gpqa_diamond,49.0,https://arxiv.org/abs/2501.12948,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek R1 paper Table 15: deepseek-r1-distill-llama-8b=49.0.
deepseek-r1-distill-llama-8b,livecodebench,42.5,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Qwen3 tech report (arxiv:2505.09388) Table 19 baseline column for deepseek-r1-distill-llama-8b: LiveCodeBench v5 42.5. Third-party Qwen self-test, matches_canonical=false."
deepseek-r1-distill-llama-8b,codeforces_rating,1205.0,https://arxiv.org/abs/2501.12948,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek R1 paper Table 15: deepseek-r1-distill-llama-8b=1205.
deepseek-r1-distill-llama-8b,aime_2025,27.8,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Qwen3 tech report (arxiv:2505.09388) Table 19 baseline column for deepseek-r1-distill-llama-8b: AIME 2025 27.8. Third-party Qwen self-test, matches_canonical=false."
deepseek-r1-distill-llama-8b,ifeval,59.0,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Qwen3 tech report (arxiv:2505.09388) Table 19 baseline column for deepseek-r1-distill-llama-8b: IFEval strict 59.0. Third-party Qwen self-test, matches_canonical=false."
deepseek-r1-distill-llama-8b,arena_hard,17.6,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Qwen3 tech report (arxiv:2505.09388) Table 19 baseline column for deepseek-r1-distill-llama-8b: Arena-Hard 17.6. Third-party Qwen self-test, matches_canonical=false."
deepseek-r1-distill-llama-70b,aime_2024,70.0,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Qwen3 tech report (arxiv:2505.09388) Table 13 baseline column for DeepSeek-R1-Distill-Llama-70B (third-party Qwen self-test, matches_canonical=false): AIME 2024 70.0"
deepseek-r1-distill-llama-70b,math_500,94.5,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Qwen3 tech report (arxiv:2505.09388) Table 13 baseline column for DeepSeek-R1-Distill-Llama-70B (third-party Qwen self-test, matches_canonical=false): MATH-500 94.5"
deepseek-r1-distill-llama-70b,gpqa_diamond,65.2,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Qwen3 tech report (arxiv:2505.09388) Table 13 baseline column for DeepSeek-R1-Distill-Llama-70B (third-party Qwen self-test, matches_canonical=false): GPQA-Diamond 65.2"
deepseek-r1-distill-llama-70b,livecodebench,54.5,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Qwen3 tech report (arxiv:2505.09388) Table 13 baseline column for DeepSeek-R1-Distill-Llama-70B (third-party Qwen self-test, matches_canonical=false): LiveCodeBench v5 54.5 [R5d: prior unverified value 57.5 from https://arxiv.org/abs/2501.12948 deleted.]"
deepseek-r1-distill-llama-70b,codeforces_rating,1633.0,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Qwen3 tech report (arxiv:2505.09388) Table 13 baseline column for DeepSeek-R1-Distill-Llama-70B (third-party Qwen self-test, matches_canonical=false): CodeForces rating 1633"
deepseek-r1-distill-llama-70b,aime_2025,56.3,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Qwen3 tech report (arxiv:2505.09388) Table 13 baseline column for DeepSeek-R1-Distill-Llama-70B (third-party Qwen self-test, matches_canonical=false): AIME 2025 56.3 [R5d: prior unverified value 70 from https://arxiv.org/html/2501.12948v1 deleted.]"
deepseek-r1-distill-llama-70b,hmmt_feb_2025,33.3,https://arxiv.org/abs/2504.21318,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4-reasoning paper Table 1 (deepseek-r1-distill-llama-70b column): hmmt_2025=33.3 (re-sourced from https://arxiv.org/abs/2504.21318, matches BP value)"
deepseek-r1-distill-llama-70b,humaneval,80.0,https://arxiv.org/html/2501.12948v1,{},,,dropped,
deepseek-v3.2,aime_2025,93.1,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: deepseek-v3.2=93.1.
deepseek-v3.2,gpqa_diamond,82.4,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: deepseek-v3.2=82.4.
deepseek-v3.2,mmlu_pro,85.0,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: deepseek-v3.2=85.0.
deepseek-v3.2,livecodebench,83.3,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: deepseek-v3.2=83.3.
deepseek-v3.2,codeforces_rating,2386.0,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: deepseek-v3.2=2386.
deepseek-v3.2,hle,25.1,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: deepseek-v3.2=25.1.
deepseek-v3.2,swe_bench_verified,73.1,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic (bash + edit)""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: deepseek-v3.2=73.1.
deepseek-v3.2,browsecomp,51.4,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: deepseek-v3.2=51.4.
deepseek-v3.2,mmmu_pro,81.0,https://www.bentoml.com/blog/the-complete-guide-to-deepseek-models-from-v3-to-r1-and-beyond,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
deepseek-v3.2,math_500,97.3,https://artificialanalysis.ai/evaluations/math-500,{},,,needs_review,"Reference URL is AA math-500 eval page, but deepseek-v3-2 is NOT present on that page; AA shows math_500=null for this model. Possible ghost score (R5g-a). Cannot verify 97.3% from AA source."
deepseek-v3.2,frontiermath,22.1,https://epoch.ai/benchmarks/frontiermath,"{""effort"":""default"",""harness"":""Fireworks API"",""mode"":""thinking"",""tools"":""none""}",True,leaderboard,verified,"epoch.ai FrontierMath-2025-02-28-Private: fireworks/deepseek-v3p2 (DeepSeek-V3.2 Thinking; Fireworks) = 22.1%. Canonical thinking mode matches. Note: epoch ran on Fireworks due to high API error rate from DeepSeek API, but results confirmed equivalent. BP had 8.0."
deepseek-v3.2,simpleqa,35.0,https://artificialanalysis.ai/models/deepseek-v3-2,{},,,needs_review,Reference URL is AA model page (deepseek-v3-2). AA does not publish SimpleQA scores; field is not present in AA model data for any model. Cannot verify from AA source. Original source needs identification.
deepseek-v3.2,terminal_bench,46.4,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""Claude Code framework"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: deepseek-v3.2=46.4.
deepseek-v3.2,aime_2024,93.0,https://introl.com/blog/deepseek-v3-2-open-source-ai-cost-advantage,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
deepseek-v3.2,brumo_2025,96.67,https://matharena.ai/?comp=brumo--brumo_2025,"""thinking:default""",True,leaderboard,verified,
deepseek-v3.2,cmimc_2025,83.75,https://matharena.ai/?comp=cmimc--cmimc_2025,"""thinking:default""",True,leaderboard,verified,
deepseek-v3.2,hmmt_nov_2025,90.2,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: deepseek-v3.2=90.2.
deepseek-v3.2,humaneval,90.0,https://introl.com/blog/deepseek-v3-2-open-source-ai-cost-advantage,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
deepseek-v3.2,ifeval,89.0,https://introl.com/blog/deepseek-v3-2-open-source-ai-cost-advantage,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
deepseek-v3.2,livebench,63.1,https://livebench.ai/,"{""effort"":""default"",""mode"":""thinking"",""tools"":""n/a""}",True,leaderboard,verified,livebench.ai table_2026_01_08 (latest as of 2026-04-28): deepseek-v3.2-thinking avg=63.1. Canonical thinking mode matches. BP had 62.2.
deepseek-v3.2,matharena_apex_2025,2.08,https://matharena.ai/?comp=apex--apex_2025,"""thinking:default""",True,leaderboard,verified,
deepseek-v3.2,arc_agi_1,57.0,https://arcprize.org/leaderboard,"{""effort"":""n/a"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""non-thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""Base LLM"",""tools"":""none""}",True,leaderboard,verified,arcprize.org leaderboard audit: Deepseek V3.2 Base LLM on leaderboard
deepseek-v3.2,arc_agi_2,4.0,https://arcprize.org/leaderboard,"{""effort"":""n/a"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""non-thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""Base LLM"",""tools"":""none""}",True,leaderboard,verified,arcprize.org leaderboard audit: Deepseek V3.2 Base LLM on leaderboard
deepseek-v3.2,mmlu,90.5,https://introl.com/blog/deepseek-v3-2-open-source-ai-cost-advantage,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
deepseek-v3.2,smt_2025,87.74,https://matharena.ai/?comp=smt--smt_2025,"""thinking:default""",True,leaderboard,verified,
deepseek-v3.2-speciale,aime_2025,96.0,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,tech_report,verified,DS V3.2 tech report Table 3: V3.2-Speciale=96.0.
deepseek-v3.2-speciale,gpqa_diamond,85.7,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,tech_report,verified,DS V3.2 tech report Table 3: V3.2-Speciale=85.7.
deepseek-v3.2-speciale,livecodebench,88.7,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,tech_report,verified,DS V3.2 tech report Table 3: V3.2-Speciale=88.7.
deepseek-v3.2-speciale,codeforces_rating,2701.0,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,tech_report,verified,DS V3.2 tech report Table 3: V3.2-Speciale=2701.
deepseek-v3.2-speciale,hmmt_feb_2025,99.2,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,tech_report,verified,DS V3.2 tech report Table 3: V3.2-Speciale=99.2.
deepseek-v3.2-speciale,hle,30.6,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,tech_report,verified,DS V3.2 tech report Table 3: V3.2-Speciale=30.6.
deepseek-v3.2-speciale,swe_bench_verified,76.0,https://arxiv.org/html/2512.02556v1,{},,,flagged," | FLAGGED: arxiv 2512.02556v1 is DeepSeek V3.2 paper; paper Table 3 has no SWE-bench for Speciale; 76.0 in paper = GPT-5 IMOAnswerBench score, not Speciale SWE-bench"
deepseek-v3.2-speciale,aime_2024,96.0,https://medium.com/@leucopsis/deepseek-v3-2-speciale-open-weights-reasoning-close-to-the-frontier-models-d43cd5da22d9,{},,,dropped,"| R5h: medium.com blog post; no primary technical report citation; needs human review DROPPED (R5h): random third-party blog/aggregator, no primary trail"
deepseek-v3.2-speciale,brumo_2025,99.17,https://matharena.ai/?comp=brumo--brumo_2025,{},,,verified, | Verified: matharena.ai shows 99.17%
deepseek-v3.2-speciale,cmimc_2025,94.38,https://matharena.ai/?comp=cmimc--cmimc_2025,{},,,verified, | Verified: matharena.ai shows 94.38%
deepseek-v3.2-speciale,hmmt_nov_2025,94.4,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,tech_report,verified,DS V3.2 tech report Table 3: V3.2-Speciale=94.4.
deepseek-v3.2-speciale,humaneval,91.5,https://llm-stats.com/models/deepseek-v3.2-speciale,{},,,dropped,"| R5h: llm-stats.com is an aggregator; no primary citation; needs human review DROPPED (R5h): random third-party blog/aggregator, no primary trail"
deepseek-v3.2-speciale,ifeval,88.0,https://llm-stats.com/models/deepseek-v3.2-speciale,{},,,dropped,"| R5h: llm-stats.com is an aggregator; no primary citation; needs human review DROPPED (R5h): random third-party blog/aggregator, no primary trail"
deepseek-v3.2-speciale,imo_2025,83.3,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""context"":""128K"",""effort"":""max"",""harness"":""official"",""judge"":""official"",""mode"":""thinking"",""notes"":""Used generate-verify-refine loop per Shao et al. 2025; 35/42=83.3%"",""prompt_style"":""generate-verify-refine"",""sampling"":""competition"",""temperature"":""1.0"",""tools"":""none""}",True,tech_report,verified,"DS V3.2 tech report Table 4: Speciale IMO 2025 = 35/42 pts = 83.3% (Gold medal). Generate-verify-refine loop, max 128K tokens, no tools."
deepseek-v3.2-speciale,math_500,98.0,https://llm-stats.com/models/deepseek-v3.2-speciale,{},,,dropped,"| R5h: llm-stats.com is an aggregator; no primary citation; needs human review DROPPED (R5h): random third-party blog/aggregator, no primary trail"
deepseek-v3.2-speciale,matharena_apex_2025,9.38,https://matharena.ai/?comp=apex--apex_2025,{},,,verified, | Verified: matharena.ai shows 9.38%
deepseek-v3.2-speciale,mmlu_pro,87.5,https://llm-stats.com/models/deepseek-v3.2-speciale,{},,,dropped,"| R5h: llm-stats.com is an aggregator; no primary citation; needs human review DROPPED (R5h): random third-party blog/aggregator, no primary trail"
deepseek-v3.2-speciale,smt_2025,89.15,https://matharena.ai/?comp=smt--smt_2025,{},,,verified, | Verified: matharena.ai shows 89.15%
deepseek-v4-pro,mmlu_pro,87.5,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-pro (Max mode) = 87.5.
deepseek-v4-pro,gpqa_diamond,90.1,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-pro (Max mode) = 90.1.
deepseek-v4-pro,livecodebench,93.5,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-pro (Max mode) = 93.5.
deepseek-v4-pro,codeforces_rating,3206.0,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-pro (Max mode) = 3206.
deepseek-v4-pro,swe_bench_verified,80.6,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-pro (Max mode) = 80.6.
deepseek-v4-pro,swe_bench_pro,55.4,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-pro (Max mode) = 55.4.
deepseek-v4-pro,terminal_bench,67.9,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-pro (Max mode) = 67.9.
deepseek-v4-pro,browsecomp,83.4,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-pro (Max mode) = 83.4.
deepseek-v4-pro,matharena_apex_2025,38.3,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,"DeepSeek-V4-Pro card Table 2 (cross-mode): V4-Pro Max Apex=38.3, confirmed in frontier table."
deepseek-v4-pro,gdpval_aa_elo,1554.0,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-pro (Max mode) = 1554.
deepseek-v4-flash,mmlu_pro,86.2,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-flash (Max mode) = 86.2.
deepseek-v4-flash,gpqa_diamond,88.1,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-flash (Max mode) = 88.1.
deepseek-v4-flash,hle,34.8,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-flash (Max mode) = 34.8.
deepseek-v4-flash,livecodebench,91.6,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-flash (Max mode) = 91.6.
deepseek-v4-flash,codeforces_rating,3052.0,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-flash (Max mode) = 3052.
deepseek-v4-flash,swe_bench_verified,79.0,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-flash (Max mode) = 79.0.
deepseek-v4-flash,swe_bench_pro,52.6,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-flash (Max mode) = 52.6.
deepseek-v4-flash,terminal_bench,56.9,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-flash (Max mode) = 56.9.
deepseek-v4-flash,browsecomp,73.2,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-flash (Max mode) = 73.2.
deepseek-v4-flash,matharena_apex_2025,33.0,https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek-V4-Flash card Table 2 (cross-mode comparison): V4-Flash Max Apex=33.0.
deepseek-v4-flash,gdpval_aa_elo,1395.0,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-flash (Max mode) = 1395.
qwen3-235b,aime_2025,81.5,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report (arxiv:2505.09388) Table 11 Thinking column: AIME 2025 81.5
qwen3-235b,gpqa_diamond,71.1,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report (arxiv:2505.09388) Table 11 Thinking column: GPQA Diamond 71.1
qwen3-235b,livecodebench,70.7,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report (arxiv:2505.09388) Table 11 Thinking column: LiveCodeBench v5 (2024.10-2025.02) 70.7
qwen3-235b,codeforces_rating,2056.0,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report (arxiv:2505.09388) Table 11 Thinking column: CodeForces rating 2056 (98.2 percentile)
qwen3-235b,arena_hard,95.6,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report (arxiv:2505.09388) Table 11 Thinking column: Arena-Hard 95.6
qwen3-235b,aime_2024,85.7,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report (arxiv:2505.09388) Table 11 Thinking column: AIME 2024 85.7
qwen3-235b,swe_bench_verified,69.6,https://qwenlm.github.io/blog/qwen3/,"""Qwen3 blog performance table, thinking mode""",True,official_blog,verified,Confirmed via multiple independent web sources (69.6%)
qwen3-235b,hmmt_feb_2025,62.5,https://huggingface.co/deepseek-ai/DeepSeek-R1-0528,{},True,model_card,verified,DeepSeek-R1-0528 HF card comparison table shows Qwen3-235B-A22B HMMT Feb 25=62.5.
qwen3-235b,hle,15.43,https://scale.com/leaderboard/humanitys_last_exam,{},,,dropped," [Dropped: Model not found in current Scale.com HLE leaderboard (50 entries, 2026-04-29 scrape). Ghost cell per R5g-a-ghost-cell.]"
qwen3-235b,mmlu_pro,79.8,https://arxiv.org/abs/2505.09388,{},,,dropped," | DROPPED: BP value 79.8 from arxiv:2505.09388 is from Non-thinking variant (Table 12), but qwen3-235b canonical = Thinking. MMLU-Pro not reported in Table 11. Cell value is wrong-variant."
qwen3-235b,math_500,98.0,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report (arxiv:2505.09388) Table 11 Thinking: MATH-500 98.0. [R5d: prior unverified value 98.2 from https://arxiv.org/abs/2505.09388 deleted.]
qwen3-235b,chatbot_arena_elo,1410.0,https://lmarena.ai/,{},,,needs_review,
qwen3-235b,gsm8k,94.39,https://arxiv.org/abs/2505.09388,{},,,dropped," | DROPPED: BP value 94.39 from arxiv:2505.09388 Table 3 (Base column). GSM8K is not reported in Table 11 (Thinking). Cell value is wrong-variant (Base, not Thinking)."
qwen3-235b,humaneval,90.0,https://qwenlm.github.io/blog/qwen3/,"""Qwen3 blog performance table, thinking mode""",True,official_blog,verified,Confirmed via official Qwen3 blog and web search (90.0%)
qwen3-235b,ifeval,83.4,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report (arxiv:2505.09388) Table 11 Thinking: IFEval strict prompt 83.4. [R5d: prior unverified value 87.8 from https://huggingface.co/Qwen/Qwen3-235B-A22B-Thinking-2507 deleted.]
qwen3-235b,mmlu,87.81,https://arxiv.org/abs/2505.09388,{},,,dropped," | DROPPED: BP value 87.81 from arxiv:2505.09388 Table 3 (Base column, MMLU 5-shot), but qwen3-235b is_reasoning=true → canonical = Thinking variant. MMLU is not reported in Table 11 (Thinking). Cell value is wrong-variant."
qwen3-235b,simpleqa,13.2,https://arxiv.org/abs/2507.20534,"{""context"":""8192 max output tokens"",""effort"":""default"",""harness"":""Kimi team evaluation (official API)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unified (Kimi team setting)"",""tools"":""none (benchmark-specified)""}",False,tech_report,verified,"Via third-party self-test by Moonshot AI (Kimi K2 tech report, Table 3, p.16). SimpleQA Correct, non-thinking (vendor-recommended no-thinking regime), 8192 max output tokens."
qwen3-235b,aa_lcr,67.0,https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning,"{""effort"":""default (reasoning variant)"",""harness"":""AA standard evaluation"",""mode"":""thinking"",""notes"":""Per AA: qwen3-235b-a22b-instruct-2507-reasoning lcr=67%"",""sampling"":""pass@1""}",,third_party,verified,ArtificialAnalysis aggregator; used as fallback when no primary source available (R5d-aggregator-fallback). AA LCR eval page: qwen3-235b-a22b-instruct-2507-reasoning lcr=0.670 → 67.0%. AA slug: qwen3-235b-a22b-instruct-2507-reasoning.
qwen3-235b,arc_agi_1,11.0,https://arcprize.org/leaderboard,"{""effort"":""n/a"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""non-thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""Base LLM"",""tools"":""none""}",False,leaderboard,verified,arcprize.org leaderboard audit: Qwen3-235b-a22b Instruct (25/07) on leaderboard as Base LLM (non-thinking); canonical is thinking
qwen3-235b,arc_agi_2,1.3,https://arcprize.org/leaderboard,"{""effort"":""n/a"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""non-thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""Base LLM"",""tools"":""none""}",False,leaderboard,verified,arcprize.org leaderboard audit: Qwen3-235b-a22b Instruct (25/07) on leaderboard as Base LLM (non-thinking); canonical is thinking
qwen3-235b,matharena_apex_2025,5.21,https://matharena.ai/?comp=apex--apex_2025,"""thinking:default""",True,leaderboard,verified,
qwen3-235b,terminal_bench_1,6.6,https://www.tbench.ai/leaderboard/terminal-bench/1.0,{},,leaderboard,verified,"Confirmed 6.6% on tbench.ai Terminal-Bench 1.0 leaderboard (Terminus 1 scaffold, 2025-05-15)."
qwen3-32b,gpqa_diamond,68.4,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Table 13/15/17/19 Thinking: GPQA-Diamond 68.4
qwen3-32b,aime_2025,72.9,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: AIME 2025 72.9. [R5d: prior unverified value 76.67 from https://arxiv.org/abs/2505.09388 deleted.]
qwen3-32b,livecodebench,65.7,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: LiveCodeBench v5 65.7. [R5d: prior unverified value 63.0 from https://qwenlm.github.io/blog/qwen3/ deleted.]
qwen3-32b,arena_hard,93.8,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: Arena-Hard 93.8 (Qwen self-report; differs from arena-hard-auto github). [R5d: prior unverified value 53.3 from https://github.com/lmarena/arena-hard-auto deleted.]
qwen3-32b,aime_2024,81.4,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Table 13/15/17/19 Thinking: AIME 2024 81.4
qwen3-32b,codeforces_rating,1977.0,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: CodeForces 1977 (97.7%). [R5d: prior unverified value 2020 from https://qwenlm.github.io/blog/qwen3/ deleted.]
qwen3-32b,gsm8k,93.4,https://arxiv.org/abs/2505.09388,{},,,dropped, | DROPPED: BP value sourced from Base model column (Table 3-8) but model is_reasoning=true → canonical = Thinking variant. Wrong-variant cell.
qwen3-32b,humaneval,85.0,https://qwenlm.github.io/blog/qwen3/,"""Qwen3 blog performance table, thinking mode""",True,official_blog,verified,Confirmed via official Qwen3 blog (85.0%)
qwen3-32b,ifeval,85.0,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: IFEval strict prompt 85.0. [R5d: prior unverified value 83.0 from https://arxiv.org/abs/2505.09388 deleted.]
qwen3-32b,math_500,97.2,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: MATH-500 97.2. [R5d: prior unverified value 96.0 from https://qwenlm.github.io/blog/qwen3/ deleted.]
qwen3-32b,mmlu,83.61,https://arxiv.org/abs/2505.09388,{},,,dropped, | DROPPED: BP value sourced from Base model column (Table 3-8) but model is_reasoning=true → canonical = Thinking variant. Wrong-variant cell.
qwen3-32b,mmlu_pro,70.0,https://arxiv.org/abs/2505.09388,{},,,dropped, | DROPPED: BP value from Non-thinking column (Table 12/14/16/18/20). Canonical = Thinking. Wrong-variant.
qwen3-32b,ifbench,37.3,https://github.com/allenai/IFBench,"""IFBench score, 0-shot""",True,leaderboard,verified,
qwen3-32b,terminal_bench_1,15.5,https://www.tbench.ai/leaderboard/terminal-bench/1.0,{},,leaderboard,verified,"Confirmed 15.5% on tbench.ai Terminal-Bench 1.0 leaderboard (TerminalAgent scaffold, 2025-07-31)."
qwen3-4b,aime_2025,65.6,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Table 13/15/17/19 Thinking: AIME 2025 65.6
qwen3-4b,math_500,97.0,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Table 13/15/17/19 Thinking: MATH-500 97.0
qwen3-4b,codeforces_rating,1671.0,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Table 13/15/17/19 Thinking: CodeForces 1671 (92.8%)
qwen3-4b,arena_hard,76.6,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: Arena-Hard 76.6. [R5d: prior unverified value 13.2 from https://github.com/lmarena/arena-hard-auto deleted.]
qwen3-4b,gpqa_diamond,55.9,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: GPQA-Diamond 55.9. [R5d: prior unverified value 51.2 from https://arxiv.org/abs/2505.09388 deleted.]
qwen3-4b,aime_2024,73.8,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Table 13/15/17/19 Thinking: AIME 2024 73.8
qwen3-4b,gsm8k,87.79,https://arxiv.org/abs/2505.09388,{},,,dropped, | DROPPED: BP value sourced from Base model column (Table 3-8) but model is_reasoning=true → canonical = Thinking variant. Wrong-variant cell.
qwen3-4b,hmmt_feb_2025,55.5,https://huggingface.co/Qwen/Qwen3-4B-Thinking-2507,{},,,dropped," | DROPPED: BP value from Qwen3-X-Thinking-2507 (post-July fine-tune), but BP model release_date=2025-05-15. Wrong-variant (later checkpoint)."
qwen3-4b,humaneval,75.0,https://qwenlm.github.io/blog/qwen3/,{},,,needs_review,Blog tables are images; web search gives approximate ~61-63% vs claimed 75.0; cannot confirm exact value. R5g(c)
qwen3-4b,ifeval,81.9,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: IFEval 81.9 (May 2025 release; replaces 87.4 from Thinking-2507). [R5d: prior unverified value 87.4 from https://huggingface.co/Qwen/Qwen3-4B-Thinking-2507 deleted.]
qwen3-4b,livecodebench,54.2,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: LiveCodeBench v5 54.2 (replaces 55.2 from Thinking-2507). [R5d: prior unverified value 55.2 from https://huggingface.co/Qwen/Qwen3-4B-Thinking-2507 deleted.]
qwen3-4b,mmlu,72.99,https://arxiv.org/abs/2505.09388,{},,,dropped, | DROPPED: BP value sourced from Base model column (Table 3-8) but model is_reasoning=true → canonical = Thinking variant. Wrong-variant cell.
qwen3-4b,mmlu_pro,74.0,https://huggingface.co/Qwen/Qwen3-4B-Thinking-2507,{},,,dropped," | DROPPED: BP value from Qwen3-X-Thinking-2507 (post-July fine-tune), but BP model release_date=2025-05-15. Wrong-variant (later checkpoint)."
qwen3-0.6b,math_500,77.6,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Table 13/15/17/19 Thinking: MATH-500 77.6
qwen3-0.6b,mmlu,52.81,https://arxiv.org/abs/2505.09388,{},,,dropped, | DROPPED: BP value sourced from Base model column (Table 3-8) but model is_reasoning=true → canonical = Thinking variant. Wrong-variant cell.
qwen3-0.6b,mmlu_pro,24.74,https://arxiv.org/abs/2505.09388,{},,,dropped, | DROPPED: BP value sourced from Base model column (Table 3-8) but model is_reasoning=true → canonical = Thinking variant. Wrong-variant cell.
qwen3-0.6b,gpqa_diamond,27.9,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Table 13/15/17/19 Thinking: GPQA-Diamond 27.9
qwen3-0.6b,aime_2024,10.7,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Table 13/15/17/19 Thinking: AIME 2024 10.7
qwen3-0.6b,aime_2025,15.1,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Table 13/15/17/19 Thinking: AIME 2025 15.1
qwen3-0.6b,ifeval,59.2,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Table 13/15/17/19 Thinking: IFEval 59.2
qwen3-0.6b,arena_hard,8.5,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Table 13/15/17/19 Thinking: Arena-Hard 8.5
qwen3-0.6b,livecodebench,12.3,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Table 13/15/17/19 Thinking: LiveCodeBench v5 12.3
qwen3-0.6b,codeforces_rating,800.0,https://qwenlm.github.io/blog/qwen3/,{},,,needs_review,Web search: 'no official Codeforces ELO given for Qwen3-0.6B'; value 800 not confirmed in any accessible text source. R5g(c)
qwen3-0.6b,gsm8k,59.59,https://arxiv.org/abs/2505.09388,{},,,dropped, | DROPPED: BP value sourced from Base model column (Table 3-8) but model is_reasoning=true → canonical = Thinking variant. Wrong-variant cell.
qwen3-0.6b,humaneval,45.0,https://arxiv.org/abs/2505.09388,{},,,dropped,
qwen3-1.7b,mmlu,61.0,https://arxiv.org/abs/2505.09388,{},,,dropped, | DROPPED: BP value sourced from Base model column (Table 3-8) but model is_reasoning=true → canonical = Thinking variant. Wrong-variant cell.
qwen3-1.7b,mmlu_pro,56.68,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""default"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: qwen3-1.7b=56.68. (third-party self-test by Liquid AI)
qwen3-1.7b,math_500,93.4,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Table 13/15/17/19 Thinking: MATH-500 93.4
qwen3-1.7b,gpqa_diamond,40.1,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Table 13/15/17/19 Thinking: GPQA-Diamond 40.1
qwen3-1.7b,aime_2024,48.3,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Table 13/15/17/19 Thinking: AIME 2024 48.3
qwen3-1.7b,aime_2025,36.8,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Table 13/15/17/19 Thinking: AIME 2025 36.8
qwen3-1.7b,ifeval,72.5,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Table 13/15/17/19 Thinking: IFEval 72.5
qwen3-1.7b,arena_hard,43.1,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Table 13/15/17/19 Thinking: Arena-Hard 43.1
qwen3-1.7b,livecodebench,33.2,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Table 13/15/17/19 Thinking: LiveCodeBench v5 33.2
qwen3-1.7b,codeforces_rating,1200.0,https://qwenlm.github.io/blog/qwen3/,{},,,needs_review,Web search: 'no official Codeforces ELO given for Qwen3-1.7B'; value 1200 not confirmed in any accessible text source. R5g(c)
qwen3-1.7b,gsm8k,85.6,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""default"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: qwen3-1.7b=85.6. (third-party self-test by Liquid AI)
qwen3-1.7b,humaneval,60.0,https://arxiv.org/abs/2505.09388,{},,,dropped,
qwen3-8b,gpqa_diamond,62.0,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: GPQA-Diamond 62.0. [R5d: prior unverified value 63.3 from https://arxiv.org/abs/2505.09388 deleted.]
qwen3-8b,mmlu,76.89,https://arxiv.org/abs/2505.09388,{},,,dropped, | DROPPED: BP value sourced from Base model column (Table 3-8) but model is_reasoning=true → canonical = Thinking variant. Wrong-variant cell.
qwen3-8b,mmlu_pro,56.73,https://arxiv.org/abs/2505.09388,{},,,dropped, | DROPPED: BP value sourced from Base model column (Table 3-8) but model is_reasoning=true → canonical = Thinking variant. Wrong-variant cell.
qwen3-8b,aime_2024,76.0,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Table 13/15/17/19 Thinking: AIME 2024 76.0
qwen3-8b,aime_2025,67.3,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: AIME 2025 67.3. [R5d: prior unverified value 67.3 from https://huggingface.co/deepseek-ai/DeepSeek-R1-0528 deleted.]
qwen3-8b,arena_hard,85.8,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: Arena-Hard 85.8. [R5d: prior unverified value 76.0 from https://qwenlm.github.io/blog/qwen3/ deleted.]
qwen3-8b,codeforces_rating,1785.0,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: CodeForces 1785 (95.6%). [R5d: prior unverified value 1850 from https://qwenlm.github.io/blog/qwen3/ deleted.]
qwen3-8b,gsm8k,89.84,https://arxiv.org/abs/2505.09388,{},,,dropped, | DROPPED: BP value sourced from Base model column (Table 3-8) but model is_reasoning=true → canonical = Thinking variant. Wrong-variant cell.
qwen3-8b,humaneval,82.0,https://qwenlm.github.io/blog/qwen3/,"""Qwen3 blog performance table, thinking mode""",True,official_blog,verified,Confirmed via official Qwen3 blog (82.0%)
qwen3-8b,ifeval,85.0,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: IFEval 85.0. [R5d: prior unverified value 80.0 from https://arxiv.org/abs/2505.09388 deleted.]
qwen3-8b,livecodebench,57.5,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: LiveCodeBench v5 57.5. [R5d: prior unverified value 48.0 from https://qwenlm.github.io/blog/qwen3/ deleted.]
qwen3-8b,math_500,97.4,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: MATH-500 97.4. [R5d: prior unverified value 94.0 from https://qwenlm.github.io/blog/qwen3/ deleted.]
qwen3-8b,ifbench,35.0,https://github.com/allenai/IFBench,"""IFBench score, 0-shot""",True,leaderboard,verified,
qwen3-14b,mmlu_pro,61.03,https://arxiv.org/abs/2505.09388,{},,,dropped, | DROPPED: BP value sourced from Base model column (Table 3-8) but model is_reasoning=true → canonical = Thinking variant. Wrong-variant cell.
qwen3-14b,mmlu,81.05,https://arxiv.org/abs/2505.09388,{},,,dropped, | DROPPED: BP value sourced from Base model column (Table 3-8) but model is_reasoning=true → canonical = Thinking variant. Wrong-variant cell.
qwen3-14b,arena_hard,91.7,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: Arena-Hard 91.7. [R5d: prior unverified value 85.5 from https://arxiv.org/abs/2505.09388 deleted.]
qwen3-14b,aime_2024,79.3,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: AIME 2024 79.3. [R5d: prior unverified value 75.0 from https://arxiv.org/abs/2505.09388 deleted.]
qwen3-14b,aime_2025,70.4,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: AIME 2025 70.4. [R5d: prior unverified value 72.0 from https://arxiv.org/abs/2505.09388 deleted.]
qwen3-14b,codeforces_rating,1766.0,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: CodeForces 1766 (95.3%). [R5d: prior unverified value 1900 from https://qwenlm.github.io/blog/qwen3/ deleted.]
qwen3-14b,gpqa_diamond,64.0,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Table 13/15/17/19 Thinking: GPQA-Diamond 64.0
qwen3-14b,gsm8k,92.49,https://arxiv.org/abs/2505.09388,{},,,dropped, | DROPPED: BP value sourced from Base model column (Table 3-8) but model is_reasoning=true → canonical = Thinking variant. Wrong-variant cell.
qwen3-14b,humaneval,85.0,https://qwenlm.github.io/blog/qwen3/,{},,,needs_review,Blog tables are images; value 85.0 not confirmed with exact precision from accessible text sources. R5g(c)
qwen3-14b,livecodebench,63.5,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: LiveCodeBench v5 63.5. [R5d: prior unverified value 55.0 from https://qwenlm.github.io/blog/qwen3/ deleted.]
qwen3-14b,math_500,96.8,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: MATH-500 96.8. [R5d: prior unverified value 95.0 from https://qwenlm.github.io/blog/qwen3/ deleted.]
qwen3-30b-a3b,gpqa_diamond,65.8,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Table 13/15/17/19 Thinking: GPQA-Diamond 65.8
qwen3-30b-a3b,codeforces_rating,1974.0,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Table 13/15/17/19 Thinking: CodeForces 1974 (97.7%)
qwen3-30b-a3b,arena_hard,91.0,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Table 13/15/17/19 Thinking: Arena-Hard 91.0
qwen3-30b-a3b,aime_2024,80.4,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: AIME 2024 80.4. [R5d: prior unverified value 78.0 from https://qwenlm.github.io/blog/qwen3/ deleted.]
qwen3-30b-a3b,aime_2025,70.9,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: AIME 2025 70.9 (May 2025 release; replaces 85.0 from Thinking-2507 fine-tune). [R5d: prior unverified value 85.0 from https://huggingface.co/Qwen/Qwen3-30B-A3B-Thinking-2507 deleted.]
qwen3-30b-a3b,gsm8k,91.81,https://arxiv.org/abs/2505.09388,{},,,dropped, | DROPPED: BP value sourced from Base model column (Table 3-8) but model is_reasoning=true → canonical = Thinking variant. Wrong-variant cell.
qwen3-30b-a3b,hmmt_feb_2025,71.4,https://huggingface.co/Qwen/Qwen3-30B-A3B-Thinking-2507,{},,,dropped," | DROPPED: BP value from Qwen3-X-Thinking-2507 (post-July fine-tune), but BP model release_date=2025-05-15. Wrong-variant (later checkpoint)."
qwen3-30b-a3b,humaneval,82.0,https://qwenlm.github.io/blog/qwen3/,{},,,needs_review,Blog tables are images; value 82.0 approximately consistent but exact confirmation unavailable. R5g(c)
qwen3-30b-a3b,ifeval,86.5,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: IFEval 86.5. [R5d: prior unverified value 88.9 from https://huggingface.co/Qwen/Qwen3-30B-A3B-Thinking-2507 deleted.]
qwen3-30b-a3b,livecodebench,62.6,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: LiveCodeBench v5 62.6. [R5d: prior unverified value 66.0 from https://huggingface.co/Qwen/Qwen3-30B-A3B-Thinking-2507 deleted.]
qwen3-30b-a3b,math_500,98.0,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: MATH-500 98.0. [R5d: prior unverified value 96.0 from https://qwenlm.github.io/blog/qwen3/ deleted.]
qwen3-30b-a3b,mmlu,81.38,https://arxiv.org/abs/2505.09388,{},,,dropped, | DROPPED: BP value sourced from Base model column (Table 3-8) but model is_reasoning=true → canonical = Thinking variant. Wrong-variant cell.
qwen3-30b-a3b,mmlu_pro,80.9,https://huggingface.co/Qwen/Qwen3-30B-A3B-Thinking-2507,{},,,dropped," | DROPPED: BP value from Qwen3-X-Thinking-2507 (post-July fine-tune), but BP model release_date=2025-05-15. Wrong-variant (later checkpoint)."
qwq-32b,aime_2024,79.5,https://qwenlm.github.io/blog/qwq-32b/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,QwQ-32B blog (qwenlm.github.io/blog/qwq-32b/) + Qwen3 tech report Table 13 baseline column: AIME 2024 79.5
qwq-32b,gpqa_diamond,65.6,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,QwQ-32B blog (qwenlm.github.io/blog/qwq-32b/) + Qwen3 tech report Table 13 baseline column: GPQA-Diamond 65.6 (Qwen3 T13; supersedes preview 65.2) [R5d: prior unverified value 65.2 from https://qwenlm.github.io/blog/qwq-32b-preview/ deleted.]
qwq-32b,math_500,98.0,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,QwQ-32B blog (qwenlm.github.io/blog/qwq-32b/) + Qwen3 tech report Table 13 baseline column: MATH-500 98.0 (Qwen3 T13; supersedes preview 90.6) [R5d: prior unverified value 90.6 from https://qwenlm.github.io/blog/qwq-32b-preview/ deleted.]
qwq-32b,livecodebench,62.7,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,QwQ-32B blog (qwenlm.github.io/blog/qwq-32b/) + Qwen3 tech report Table 13 baseline column: LiveCodeBench v5 62.7 (Qwen3 tech report Table 13 baseline column; supersedes earlier preview blog 50.0) [R5d: prior unverified value 50.0 from https://qwenlm.github.io/blog/qwq-32b-preview/ deleted.]
qwq-32b,arena_hard,89.5,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,QwQ-32B blog (qwenlm.github.io/blog/qwq-32b/) + Qwen3 tech report Table 13 baseline column: Arena-Hard 89.5 (Qwen3 T13; supersedes lmarena github 60.9) [R5d: prior unverified value 60.9 from https://github.com/lmarena/arena-hard-auto deleted.]
qwq-32b,codeforces_rating,1982.0,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,QwQ-32B blog (qwenlm.github.io/blog/qwq-32b/) + Qwen3 tech report Table 13 baseline column: CodeForces rating 1982 (Qwen3 T13; supersedes 1316 from DeepSeek R1 arxiv comparison) [R5d: prior unverified value 1316 from https://arxiv.org/html/2501.12948v1 deleted.]
qwq-32b,aime_2025,69.5,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,QwQ-32B blog (qwenlm.github.io/blog/qwq-32b/) + Qwen3 tech report Table 13 baseline column: AIME 2025 69.5 (Qwen3 T13) [R5d: prior unverified value 50.0 from https://matharena.ai/models/qwen_qwq_preview deleted.]
qwq-32b,ifeval,83.9,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,QwQ-32B blog (qwenlm.github.io/blog/qwq-32b/) + Qwen3 tech report Table 13 baseline column: IFEval strict prompt 83.9 [R5d: prior unverified value 80.0 from https://qwenlm.github.io/blog/qwq-32b/ deleted.]
qwq-32b,mmlu_pro,62.0,https://qwenlm.github.io/blog/qwq-32b/,{},,,dropped, | DROPPED: BP cell URL points to qwenlm.github.io/blog/qwq-32b/ but that blog does not report MMLU-Pro; not in Qwen3 tech report Table 13 either. Ghost cell per R5g(a).
qwq-32b,bigcodebench,44.6,https://bigcode-bench.github.io/,"""BigCodeBench Complete Instruct pass@1""",True,leaderboard,verified,Source entry: QwQ-32B-Preview (2024-11-28); regular QwQ-32B has no instruct score in leaderboard
qwq-32b,hmmt_feb_2025,47.5,https://arxiv.org/abs/2504.21318,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4-reasoning paper Table 1 (qwq-32b column): hmmt_2025=47.5 (re-sourced from https://arxiv.org/abs/2504.21318, matches BP value)"
qwq-32b,humaneval,78.0,https://medium.com/towards-agi/qwq-32b-preview-benchmarks-revolutionizing-ai-reasoning-capabilities-b2014a00c208,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
qwq-32b,mmlu,88.4,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,OLMo 2 paper Table 7 Instruct: mmlu=88.4. [R5d: prior unverified value 79 from https://medium.com/towards-agi/qwq-32b-preview-benchmarks-revolutionizing-ai-reasoning-capabilities-b2014a00c208 deleted.]
qwq-32b,swe_bench_verified,35.0,https://medium.com/towards-agi/qwq-32b-preview-benchmarks-revolutionizing-ai-reasoning-capabilities-b2014a00c208,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
qwen3.5-397b,gpqa_diamond,88.4,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Qwen3.5-397B-A17B HF model card cross-model table: GPQA Diamond 88.4 (was 92.4 = GPT5.2 col). [R5d: prior unverified value 92.4 from https://huggingface.co/Qwen/Qwen3.5-397B-A17B deleted — was scraped from wrong column (likely GPT5.2 column).]
qwen3.5-397b,swe_bench_verified,76.4,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Qwen3.5-397B-A17B HF model card cross-model table: SWE-bench Verified 76.4 (was 80.0 = GPT5.2). [R5d: prior unverified value 80.0 from https://huggingface.co/Qwen/Qwen3.5-397B-A17B deleted — was scraped from wrong column (likely GPT5.2 column).]
qwen3.5-397b,mmlu_pro,87.8,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Qwen3.5-397B-A17B HF model card cross-model table: MMLU-Pro 87.8 (was 87.4 = GPT5.2). [R5d: prior unverified value 87.4 from https://huggingface.co/Qwen/Qwen3.5-397B-A17B deleted — was scraped from wrong column (likely GPT5.2 column).]
qwen3.5-397b,livecodebench,83.6,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Qwen3.5-397B-A17B HF model card cross-model table: LiveCodeBench v6 83.6 (was 87.7 = GPT5.2). [R5d: prior unverified value 87.7 from https://huggingface.co/Qwen/Qwen3.5-397B-A17B deleted — was scraped from wrong column (likely GPT5.2 column).]
qwen3.5-397b,ifeval,92.6,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Qwen3.5-397B-A17B HF model card cross-model table: IFEval 92.6 (was 94.8 = GPT5.2). [R5d: prior unverified value 94.8 from https://huggingface.co/Qwen/Qwen3.5-397B-A17B deleted — was scraped from wrong column (likely GPT5.2 column).]
qwen3.5-397b,aime_2025,83.1,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image1.png?download=true,"{""context"":""default (262K)"",""effort"":""default (maximum reasoning setting)"",""harness"":""Mistral official maximum-reasoning comparison"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default (thinking by default)"",""sampling"":""samples=16 (reported as avg@16)"",""temperature"":""default"",""tools"":""none""}",False,official_model_card,verified,"Displayed exactly: '83.1'. Research observation: medium-image1.png:1:5:score:reported. Chart footnotes: *All model evaluations were run with maximum reasoning settings. | **Sonnet 4.6 encountered higher rates of reasoning truncation due to external API restrictions than other models, which affected its performance."
qwen3.5-397b,humaneval,92.0,https://venturebeat.com/technology/alibabas-qwen-3-5-397b-a17/,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
qwen3.5-397b,math_500,98.0,https://venturebeat.com/technology/alibabas-qwen-3-5-397b-a17/,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
qwen3.5-397b,mmlu,88.6,https://automatio.ai/models/qwen3-5-397b-a17b,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
qwen3.5-397b,aime_2024,94.0,https://artificialanalysis.ai/models/qwen3-5-397b-a17b,{},,,needs_review,Reference URL is AA model page (qwen3-5-397b-a17b). AA shows aime=null for all qwen3-5 models; AIME benchmark not yet evaluated by AA for this model. Cannot verify 94 from AA source.
qwen3.5-397b,aime_2026,91.3,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Qwen3.5-397B-A17B HF model card cross-model table: AIME 2026 91.3 (was 96.7 = GPT5.2). [R5d: prior unverified value 96.7 from https://huggingface.co/Qwen/Qwen3.5-397B-A17B deleted — was scraped from wrong column (likely GPT5.2 column).]
qwen3.5-397b,codeforces_rating,2200.0,https://artificialanalysis.ai/models/qwen3-5-397b-a17b,{},,,needs_review,Reference URL is AA model page (qwen3-5-397b-a17b). AA does not publish codeforces_rating data; field is not present in AA model data. Cannot verify 2200 from AA source. Original source may be Qwen tech report.
qwen3.5-397b,hle,28.7,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Qwen3.5-397B-A17B HF model card cross-model table: HLE 28.7 (was 35.5 = GPT5.2). [R5d: prior unverified value 35.5 from https://huggingface.co/Qwen/Qwen3.5-397B-A17B deleted — was scraped from wrong column (likely GPT5.2 column).]
qwen3.5-397b,ifbench,76.5,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Qwen3.5-397B-A17B HF model card cross-model table: IFBench 76.5 (was 75.4 = GPT5.2). [R5d: prior unverified value 75.4 from https://huggingface.co/Qwen/Qwen3.5-397B-A17B deleted — was scraped from wrong column (likely GPT5.2 column).]
qwen3.5-397b,mathvision,88.6,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Qwen3.5-397B-A17B HF model card cross-model table: MathVision 88.6 (HF Vision section; supersedes 90.3 from artificialanalysis.ai). [R5d: prior unverified value 90.3 from https://artificialanalysis.ai/models/qwen3-5-397b-a17b deleted — was scraped from wrong column (likely GPT5.2 column).]
qwen3.5-397b,mmmu,85.0,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Qwen3.5-397B-A17B HF model card cross-model table: MMMU 85.0
qwen3.5-397b,simpleqa,35.0,https://artificialanalysis.ai/models/qwen3-5-397b-a17b,{},,,needs_review,Reference URL is AA model page (qwen3-5-397b-a17b). AA does not publish SimpleQA scores; field is not present in AA model data for any model. Cannot verify from AA source. Original source needs identification.
qwen3.5-397b,mmmlu,88.5,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Qwen3.5-397B-A17B HF model card cross-model table: MMMLU 88.5 (was 89.5 = GPT5.2). [R5d: prior unverified value 89.5 from https://huggingface.co/Qwen/Qwen3.5-397B-A17B deleted — was scraped from wrong column (likely GPT5.2 column).]
qwen3.5-397b,swe_bench_multilingual,69.3,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Qwen3.5-397B-A17B HF model card cross-model table: SWE-bench Multilingual 69.3 (was 72.0 = GPT5.2). [R5d: prior unverified value 72.0 from https://huggingface.co/Qwen/Qwen3.5-397B-A17B deleted — was scraped from wrong column (likely GPT5.2 column).]
qwen3.5-397b,imo_answerbench,80.9,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Qwen3.5-397B-A17B HF model card cross-model table: IMOAnswerBench 80.9 (was 86.3 = GPT5.2). [R5d: prior unverified value 86.3 from https://huggingface.co/Qwen/Qwen3.5-397B-A17B deleted — was scraped from wrong column (likely GPT5.2 column).]
qwen3.5-397b,terminal_bench,52.5,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Qwen3.5-397B-A17B HF model card cross-model table: Terminal-Bench 2 52.5 (was 54.0 = GPT5.2). [R5d: prior unverified value 54.0 from https://huggingface.co/Qwen/Qwen3.5-397B-A17B deleted — was scraped from wrong column (likely GPT5.2 column).]
qwen3.5-397b,browsecomp,69.0,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Qwen3.5-397B-A17B HF model card cross-model table: BrowseComp 69.0 simple context-folding (was 65.8 = GPT5.2). [R5d: prior unverified value 65.8 from https://huggingface.co/Qwen/Qwen3.5-397B-A17B deleted — was scraped from wrong column (likely GPT5.2 column).]
qwen3.5-397b,aa_lcr,68.7,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Qwen3.5-397B-A17B HF model card cross-model table: AA-LCR 68.7 (was 72.7 = GPT5.2). [R5d: prior unverified value 72.7 from https://huggingface.co/Qwen/Qwen3.5-397B-A17B deleted — was scraped from wrong column (likely GPT5.2 column).]
qwen3.5-397b,hmmt_feb_2025,94.8,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Qwen3.5-397B-A17B HF model card cross-model table: HMMT Feb 25 94.8 (was 99.4 = GPT5.2). [R5d: prior unverified value 99.4 from https://huggingface.co/Qwen/Qwen3.5-397B-A17B deleted — was scraped from wrong column (likely GPT5.2 column).]
qwen3.5-397b,hmmt_nov_2025,92.7,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Qwen3.5-397B-A17B HF model card cross-model table: HMMT Nov 25 92.7 (was 100.0 = GPT5.2). [R5d: prior unverified value 100.0 from https://huggingface.co/Qwen/Qwen3.5-397B-A17B deleted — was scraped from wrong column (likely GPT5.2 column).]
qwen3.6-plus,swe_bench_verified,78.8,https://qwen.ai/blog?id=qwen3.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,"Qwen3.6 blog: qwen3.6-plus=78.8 (vs Claude Opus 4.5=80.9, Kimi-K2.5=76.8, GLM5=77.8, Qwen3.5-397B=76.2)."
qwen3.6-plus,swe_bench_pro,56.6,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,GLM-5.1 blog: qwen3.6-plus=56.6.
qwen3.6-plus,terminal_bench,61.6,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5.1 blog: qwen3.6-plus=61.6.
qwen3.6-plus,hle,28.8,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5.1 blog: qwen3.6-plus=28.8.
qwen3.6-plus,mmlu_pro,88.5,https://qwen.ai/blog?id=qwen3.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"Qwen3.6 blog: qwen3.6-plus=88.5 (vs Claude Opus 4.5=89.5, Kimi-K2.5=87.1, GLM5=85.7, Qwen3.5-397B=87.8)."
qwen3.6-plus,ifbench,74.2,https://qwen.ai/blog?id=qwen3.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"Qwen3.6 blog: qwen3.6-plus=74.2 (vs Claude Opus 4.5=58.0, Kimi-K2.5=70.2, GLM5=72.3, Qwen3.5-397B=76.5)."
qwen3.6-plus,aa_lcr,68.3,https://qwen.ai/blog?id=qwen3.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"Qwen3.6 blog: qwen3.6-plus=68.3 (vs Claude Opus 4.5=74.0, Kimi-K2.5=70.0, GLM5=63.3, Qwen3.5-397B=68.7)."
qwen3.6-plus,gpqa_diamond,90.4,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5.1 blog: qwen3.6-plus=90.4.
qwen3.6-plus,aime_2026,95.1,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5.1 blog: qwen3.6-plus=95.1.
kimi-k2,gpqa_diamond,75.1,https://arxiv.org/abs/2507.20534,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg of 4-64 trials per benchmark)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,tech_report,verified,Kimi K2 tech report (arxiv:2507.20534) Table 3 Kimi-K2-Instruct column: GPQA-Diamond Avg@8 75.1
kimi-k2,swe_bench_verified,65.8,https://arxiv.org/abs/2507.20534,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg of 4-64 trials per benchmark)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,tech_report,verified,Kimi K2 tech report (arxiv:2507.20534) Table 3 Kimi-K2-Instruct column: SWE-bench Verified Agentic-Single-Attempt 65.8
kimi-k2,livecodebench,53.7,https://arxiv.org/abs/2507.20534,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg of 4-64 trials per benchmark)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,tech_report,verified,Kimi K2 tech report (arxiv:2507.20534) Table 3 Kimi-K2-Instruct column: LiveCodeBench v6 Pass@1 53.7
kimi-k2,math_500,97.4,https://arxiv.org/abs/2507.20534,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg of 4-64 trials per benchmark)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,tech_report,verified,Kimi K2 tech report (arxiv:2507.20534) Table 3 Kimi-K2-Instruct column: MATH-500 97.4
kimi-k2,aime_2025,49.5,https://arxiv.org/abs/2507.20534,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg of 4-64 trials per benchmark)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,tech_report,verified,Kimi K2 tech report (arxiv:2507.20534) Table 3 Kimi-K2-Instruct column: AIME 2025 Avg@64 49.5
kimi-k2,aime_2024,69.6,https://arxiv.org/abs/2507.20534,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg of 4-64 trials per benchmark)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,tech_report,verified,Kimi K2 tech report (arxiv:2507.20534) Table 3 Kimi-K2-Instruct column: AIME 2024 Avg@64 69.6
kimi-k2,swe_bench_pro,27.67,https://scale.com/leaderboard/swe_bench_verified,{},,,verified,
kimi-k2,arena_hard,54.5,https://arxiv.org/abs/2507.20534,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg of 4-64 trials per benchmark)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,tech_report,verified,Kimi K2 tech report (arxiv:2507.20534) Table 3 Kimi-K2-Instruct column: Arena Hard v2.0 Hard Prompt 54.5
kimi-k2,hle,4.7,https://arxiv.org/abs/2507.20534,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg of 4-64 trials per benchmark)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,tech_report,verified,Kimi K2 tech report (arxiv:2507.20534) Table 3 Kimi-K2-Instruct column: HLE 4.7
kimi-k2,humaneval,85.0,https://github.com/MoonshotAI/Kimi-K2,{},,,dropped, | DROPPED (R5g-a-ghost-cell): not in official source
kimi-k2,ifeval,89.8,https://arxiv.org/abs/2507.20534,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg of 4-64 trials per benchmark)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,tech_report,verified,Kimi K2 tech report (arxiv:2507.20534) Table 3 Kimi-K2-Instruct column: IFEval Prompt Strict 89.8
kimi-k2,mmlu,89.5,https://arxiv.org/abs/2507.20534,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg of 4-64 trials per benchmark)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,tech_report,verified,Kimi K2 tech report (arxiv:2507.20534) Table 3 Kimi-K2-Instruct column: MMLU EM 89.5
kimi-k2,mmlu_pro,81.1,https://arxiv.org/abs/2507.20534,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg of 4-64 trials per benchmark)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,tech_report,verified,Kimi K2 tech report (arxiv:2507.20534) Table 3 Kimi-K2-Instruct column: MMLU-Pro EM 81.1 [R5d: prior unverified value 87.1 from https://huggingface.co/moonshotai/Kimi-K2.5 deleted.]
kimi-k2,simpleqa,31.0,https://arxiv.org/abs/2507.20534,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg of 4-64 trials per benchmark)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,tech_report,verified,Kimi K2 tech report (arxiv:2507.20534) Table 3 Kimi-K2-Instruct column: SimpleQA Correct 31.0
kimi-k2,terminal_bench,44.5,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: kimi-k2=44.5.
kimi-k2,codeforces_rating,1780.0,https://medium.com/data-science-in-your-pocket/kimi-k2-benchmarks-explained-5b25dd6d3a3e,{},,,dropped, | DROPPED (R5h-third-party-blog: medium.com)
kimi-k2,osworld,38.0,https://arxiv.org/html/2507.20534v1,{},,,dropped, | DROPPED (R5g-a-ghost-cell): not in official source
kimi-k2-thinking,gpqa_diamond,84.5,https://moonshotai.github.io/Kimi-K2/thinking.html,"{""effort"":""default (96k thinking-token budget for HLE/AIME/HMMT/GPQA; 128k for IMO/LCB/OJ)"",""harness"":""official"",""judge"":""rule-based (HLE uses o3-mini)"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 for AIME/HMMT no-tools)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,official_blog,verified,Kimi K2 Thinking blog cross-model table: GPQA-Diamond no-tools 84.5 [R5d: prior unverified value 84.5 from https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf deleted.]
kimi-k2-thinking,swe_bench_verified,71.3,https://moonshotai.github.io/Kimi-K2/thinking.html,"{""effort"":""default (96k thinking-token budget for HLE/AIME/HMMT/GPQA; 128k for IMO/LCB/OJ)"",""harness"":""official"",""judge"":""rule-based (HLE uses o3-mini)"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 for AIME/HMMT no-tools)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,official_blog,verified,Kimi K2 Thinking blog cross-model table: SWE-bench Verified w/ tools 71.3
kimi-k2-thinking,hle,23.9,https://moonshotai.github.io/Kimi-K2/thinking.html,"{""effort"":""default (96k thinking-token budget for HLE/AIME/HMMT/GPQA; 128k for IMO/LCB/OJ)"",""harness"":""official"",""judge"":""rule-based (HLE uses o3-mini)"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 for AIME/HMMT no-tools)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,official_blog,verified,Kimi K2 Thinking blog cross-model table: HLE Text-only no tools 23.9 (canonical=tools=none). Note: BP previously had 44.9 from venturebeat = w/ tools — promoted no-tools to primary. [R5d: prior unverified value 23.9 from https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf deleted.]
kimi-k2-thinking,browsecomp,60.2,https://moonshotai.github.io/Kimi-K2/thinking.html,"{""effort"":""default (96k thinking-token budget for HLE/AIME/HMMT/GPQA; 128k for IMO/LCB/OJ)"",""harness"":""official"",""judge"":""rule-based (HLE uses o3-mini)"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 for AIME/HMMT no-tools)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,official_blog,verified,Kimi K2 Thinking blog cross-model table: BrowseComp w/ tools 60.2
kimi-k2-thinking,aime_2025,94.5,https://moonshotai.github.io/Kimi-K2/thinking.html,"{""effort"":""default (96k thinking-token budget for HLE/AIME/HMMT/GPQA; 128k for IMO/LCB/OJ)"",""harness"":""official"",""judge"":""rule-based (HLE uses o3-mini)"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 for AIME/HMMT no-tools)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,official_blog,verified,Kimi K2 Thinking blog cross-model table: AIME 2025 no-tools avg@32 94.5
kimi-k2-thinking,mmlu_pro,84.6,https://moonshotai.github.io/Kimi-K2/thinking.html,"{""effort"":""default (96k thinking-token budget for HLE/AIME/HMMT/GPQA; 128k for IMO/LCB/OJ)"",""harness"":""official"",""judge"":""rule-based (HLE uses o3-mini)"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 for AIME/HMMT no-tools)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,official_blog,verified,Kimi K2 Thinking blog cross-model table: MMLU-Pro 84.6
kimi-k2-thinking,livecodebench,83.1,https://moonshotai.github.io/Kimi-K2/thinking.html,"{""effort"":""default (96k thinking-token budget for HLE/AIME/HMMT/GPQA; 128k for IMO/LCB/OJ)"",""harness"":""official"",""judge"":""rule-based (HLE uses o3-mini)"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 for AIME/HMMT no-tools)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,official_blog,verified,Kimi K2 Thinking blog cross-model table: LiveCodeBench v6 no-tools 83.1 [R5d: prior unverified value 82.6 from https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf deleted.]
kimi-k2-thinking,arc_agi_1,12.0,https://moonshotai.github.io/Kimi-K2/thinking.html,{},,,dropped, | DROPPED (R5g-a-ghost-cell): not in official source
kimi-k2-thinking,math_500,97.0,https://venturebeat.com/ai/moonshots-kimi-k2-thinking/,{},,,dropped, | DROPPED (R5h-third-party-blog: venturebeat/felloai)
kimi-k2-thinking,simpleqa,35.0,https://venturebeat.com/ai/moonshots-kimi-k2-thinking/,{},,,dropped, | DROPPED (R5h-third-party-blog: venturebeat/felloai)
kimi-k2-thinking,terminal_bench,47.1,https://moonshotai.github.io/Kimi-K2/thinking.html,"{""effort"":""default (96k thinking-token budget for HLE/AIME/HMMT/GPQA; 128k for IMO/LCB/OJ)"",""harness"":""official"",""judge"":""rule-based (HLE uses o3-mini)"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 for AIME/HMMT no-tools)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,official_blog,verified,Kimi K2 Thinking blog cross-model table: Terminal-Bench w/ simulated tools (Terminus-2 + JSON parser) 47.1 [R5d: prior unverified value 35.7 from https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf deleted.]
kimi-k2-thinking,aime_2024,94.0,https://felloai.com/new-chinese-model-kimi-k2-thinking-ranks-1-in-multiple-benchmarks/,{},,,dropped, | DROPPED (R5h-third-party-blog: venturebeat/felloai)
kimi-k2-thinking,brumo_2025,93.33,https://matharena.ai/?comp=brumo--brumo_2025,"""thinking:default""",True,leaderboard,verified,
kimi-k2-thinking,cmimc_2025,91.88,https://matharena.ai/?comp=cmimc--cmimc_2025,"""thinking:default""",True,leaderboard,verified,
kimi-k2-thinking,codeforces_rating,2150.0,https://felloai.com/new-chinese-model-kimi-k2-thinking-ranks-1-in-multiple-benchmarks/,{},,,dropped, | DROPPED (R5h-third-party-blog: felloai)
kimi-k2-thinking,hmmt_nov_2025,89.2,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: kimi-k2-thinking=89.2.
kimi-k2-thinking,humaneval,92.0,https://felloai.com/new-chinese-model-kimi-k2-thinking-ranks-1-in-multiple-benchmarks/,{},,,dropped, | DROPPED (R5h-third-party-blog: felloai)
kimi-k2-thinking,matharena_apex_2025,0.0,https://matharena.ai/?comp=apex--apex_2025,"""thinking:default""",True,leaderboard,verified, | DROPPED (R5h-third-party-blog: placeholder)
kimi-k2-thinking,smt_2025,91.04,https://matharena.ai/?comp=smt--smt_2025,"""thinking:default""",True,leaderboard,verified,
kimi-k2.5,hle,30.1,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default (96k for reasoning, 64k for vision)"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 AIME/HMMT, avg@8 GPQA, avg@3 vision, avg@4 Seal/WideSearch)"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,model_card,verified,Kimi K2.5 model card (HF moonshotai/Kimi-K2.5) Evaluation table: HLE Full no-tools 30.1
kimi-k2.5,browsecomp,60.6,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default (96k for reasoning, 64k for vision)"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 AIME/HMMT, avg@8 GPQA, avg@3 vision, avg@4 Seal/WideSearch)"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,model_card,verified,Kimi K2.5 model card (HF moonshotai/Kimi-K2.5) Evaluation table: BrowseComp w/ tools 60.6
kimi-k2.5,swe_bench_verified,76.8,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default (96k for reasoning, 64k for vision)"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 AIME/HMMT, avg@8 GPQA, avg@3 vision, avg@4 Seal/WideSearch)"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,model_card,verified,Kimi K2.5 model card (HF moonshotai/Kimi-K2.5) Evaluation table: SWE-Bench Verified 76.8
kimi-k2.5,mmmu_pro,78.5,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default (96k for reasoning, 64k for vision)"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 AIME/HMMT, avg@8 GPQA, avg@3 vision, avg@4 Seal/WideSearch)"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,model_card,verified,Kimi K2.5 model card (HF moonshotai/Kimi-K2.5) Evaluation table: MMMU-Pro 78.5
kimi-k2.5,video_mmmu,86.6,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default (96k for reasoning, 64k for vision)"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 AIME/HMMT, avg@8 GPQA, avg@3 vision, avg@4 Seal/WideSearch)"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,model_card,verified,Kimi K2.5 model card (HF moonshotai/Kimi-K2.5) Evaluation table: Video-MMMU 86.6
kimi-k2.5,aime_2025,96.1,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default (96k for reasoning, 64k for vision)"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 AIME/HMMT, avg@8 GPQA, avg@3 vision, avg@4 Seal/WideSearch)"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,model_card,verified,Kimi K2.5 model card (HF moonshotai/Kimi-K2.5) Evaluation table: AIME 2025 96.1
kimi-k2.5,gpqa_diamond,87.6,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default (96k for reasoning, 64k for vision)"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 AIME/HMMT, avg@8 GPQA, avg@3 vision, avg@4 Seal/WideSearch)"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,model_card,verified,Kimi K2.5 model card (HF moonshotai/Kimi-K2.5) Evaluation table: GPQA-Diamond 87.6
kimi-k2.5,livecodebench,85.0,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default (96k for reasoning, 64k for vision)"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 AIME/HMMT, avg@8 GPQA, avg@3 vision, avg@4 Seal/WideSearch)"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,model_card,verified,Kimi K2.5 model card (HF moonshotai/Kimi-K2.5) Evaluation table: LiveCodeBench v6 85.0
kimi-k2.5,mmlu,92.0,https://huggingface.co/moonshotai/Kimi-K2.5,{},,,dropped," | DROPPED: BP value 92.0 from HF moonshotai/Kimi-K2.5 — but K2.5 model card only reports MMLU-Pro 87.1, NOT MMLU. Source mislabel/scrape error."
kimi-k2.5,mmlu_pro,87.1,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default (96k for reasoning, 64k for vision)"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 AIME/HMMT, avg@8 GPQA, avg@3 vision, avg@4 Seal/WideSearch)"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,model_card,verified,Kimi K2.5 model card (HF moonshotai/Kimi-K2.5) Evaluation table: MMLU-Pro 87.1
kimi-k2.5,mmmu,84.3,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,model_card,verified,"Qwen3.5-397B-A17B HF model card cross-model table (Qwen self-test of 3rd-party model, mc=false): mmmu 84.3 (3rd-party Qwen self-test). [R5d: prior unverified value 84.0 from https://www.kimi.com/blog/kimi-k2-5.html deleted.]"
kimi-k2.5,terminal_bench,50.8,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default (96k for reasoning, 64k for vision)"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 AIME/HMMT, avg@8 GPQA, avg@3 vision, avg@4 Seal/WideSearch)"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,model_card,verified,Kimi K2.5 model card (HF moonshotai/Kimi-K2.5) Evaluation table: Terminal Bench 2.0 50.8
kimi-k2.5,aime_2024,96.1,https://www.kimi.com/blog/kimi-k2-5.html,{},,,dropped,
kimi-k2.5,arc_agi_1,65.3,https://arcprize.org/leaderboard,"{""effort"":""default"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""CoT"",""tools"":""none""}",True,leaderboard,verified,arcprize.org leaderboard audit: Kimi K2.5 on leaderboard
kimi-k2.5,arc_agi_2,11.8,https://arcprize.org/leaderboard,"{""effort"":""default"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""CoT"",""tools"":""none""}",True,leaderboard,verified,arcprize.org leaderboard audit: Kimi K2.5 on leaderboard
kimi-k2.5,brumo_2025,98.33,https://matharena.ai/?comp=brumo--brumo_2025,"""thinking:default""",True,leaderboard,verified,
kimi-k2.5,cmimc_2025,91.25,https://matharena.ai/?comp=cmimc--cmimc_2025,"""thinking:default""",True,leaderboard,verified,
kimi-k2.5,codeforces_rating,2350.0,https://www.kimi.com/blog/kimi-k2-5.html,{},,,dropped,
kimi-k2.5,frontiermath,28.0,https://kimi-k25.com/blog/kimi-k2-5-benchmark,{},,,dropped,
kimi-k2.5,hmmt_nov_2025,91.1,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5.1 blog: kimi-k2.5=91.1.
kimi-k2.5,humaneval,95.0,https://kimi-k25.com/blog/kimi-k2-5-benchmark,{},,,dropped,
kimi-k2.5,ifeval,93.9,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,model_card,verified,"Qwen3.5-397B-A17B HF model card cross-model table (Qwen self-test of 3rd-party model, mc=false): ifeval 93.9 (3rd-party Qwen self-test). [R5d: prior unverified value 90 from https://kimi-k25.com/blog/kimi-k2-5-benchmark deleted.]"
kimi-k2.5,math_500,98.0,https://www.kimi.com/blog/kimi-k2-5.html,{},,,dropped,
kimi-k2.5,matharena_apex_2025,8.85,https://matharena.ai/?comp=apex--apex_2025,"""thinking:default""",True,leaderboard,verified,
kimi-k2.5,mathvision,84.2,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default (96k for reasoning, 64k for vision)"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 AIME/HMMT, avg@8 GPQA, avg@3 vision, avg@4 Seal/WideSearch)"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,model_card,verified,Kimi K2.5 model card (HF moonshotai/Kimi-K2.5) Evaluation table: MathVision 84.2
kimi-k2.5,osworld,63.3,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.6 model card: kimi-k2.5=63.3.
kimi-k2.5,simpleqa,45.0,https://kimi-k25.com/blog/kimi-k2-5-benchmark,{},,,dropped,
kimi-k2.5,smt_2025,90.57,https://matharena.ai/?comp=smt--smt_2025,"""thinking:default""",True,leaderboard,verified,
kimi-k2.5,swe_bench_pro,50.7,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default (96k for reasoning, 64k for vision)"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 AIME/HMMT, avg@8 GPQA, avg@3 vision, avg@4 Seal/WideSearch)"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,model_card,verified,Kimi K2.5 model card (HF moonshotai/Kimi-K2.5) Evaluation table: SWE-Bench Pro 50.7
kimi-k2.5,swe_bench_multilingual,73.0,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default (96k for reasoning, 64k for vision)"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 AIME/HMMT, avg@8 GPQA, avg@3 vision, avg@4 Seal/WideSearch)"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,model_card,verified,Kimi K2.5 model card (HF moonshotai/Kimi-K2.5) Evaluation table: SWE-Bench Multilingual 73.0
kimi-k2.5,charxiv_reasoning,77.5,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default (96k for reasoning, 64k for vision)"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 AIME/HMMT, avg@8 GPQA, avg@3 vision, avg@4 Seal/WideSearch)"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,model_card,verified,Kimi K2.5 model card (HF moonshotai/Kimi-K2.5) Evaluation table: CharXiv RQ 77.5
kimi-k2.5,aa_lcr,70.0,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default (96k for reasoning, 64k for vision)"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 AIME/HMMT, avg@8 GPQA, avg@3 vision, avg@4 Seal/WideSearch)"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,model_card,verified,Kimi K2.5 model card (HF moonshotai/Kimi-K2.5) Evaluation table: AA-LCR 70.0
kimi-k2.5,scicode,48.7,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default (96k for reasoning, 64k for vision)"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 AIME/HMMT, avg@8 GPQA, avg@3 vision, avg@4 Seal/WideSearch)"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,model_card,verified,Kimi K2.5 model card (HF moonshotai/Kimi-K2.5) Evaluation table: SciCode 48.7
kimi-k2.6,hle,36.4,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): kimi-k2.6 (per their own canonical config) = 36.4.
kimi-k2.6,aime_2026,96.4,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""max (98304 generation tokens)"",""harness"":""official (in-house SWE-agent-derived for SWE-Bench)"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@10 coding, avg@3 vision)"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,model_card,verified,Kimi K2.6 model card (HF moonshotai/Kimi-K2.6) Evaluation table: AIME 2026 96.4
kimi-k2.6,gpqa_diamond,90.5,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): kimi-k2.6 (per their own canonical config) = 90.5.
kimi-k2.6,terminal_bench,66.7,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): kimi-k2.6 (per their own canonical config) = 66.7.
kimi-k2.6,swe_bench_pro,58.6,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): kimi-k2.6 (per their own canonical config) = 58.6.
kimi-k2.6,swe_bench_verified,80.2,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): kimi-k2.6 (per their own canonical config) = 80.2.
kimi-k2.6,livecodebench,89.6,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): kimi-k2.6 (per their own canonical config) = 89.6.
kimi-k2.6,browsecomp,83.2,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): kimi-k2.6 (per their own canonical config) = 83.2.
kimi-k2.6,mmmu_pro,79.4,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""max (98304 generation tokens)"",""harness"":""official (in-house SWE-agent-derived for SWE-Bench)"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@10 coding, avg@3 vision)"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,model_card,verified,Kimi K2.6 model card (HF moonshotai/Kimi-K2.6) Evaluation table: MMMU-Pro 79.4
kimi-k2.6,scicode,52.2,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""max (98304 generation tokens)"",""harness"":""official (in-house SWE-agent-derived for SWE-Bench)"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@10 coding, avg@3 vision)"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,model_card,verified,Kimi K2.6 model card (HF moonshotai/Kimi-K2.6) Evaluation table: SciCode 52.2
kimi-k2.6,mathvision,87.4,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""max (98304 generation tokens)"",""harness"":""official (in-house SWE-agent-derived for SWE-Bench)"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@10 coding, avg@3 vision)"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,model_card,verified,Kimi K2.6 model card (HF moonshotai/Kimi-K2.6) Evaluation table: MathVision 87.4
glm-4.6,swe_bench_verified,68.0,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,GLM-4.7 blog: glm-4.6=68.0.
glm-4.6,gpqa_diamond,81.0,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-4.7 blog: glm-4.6=81.0.
glm-4.6,hle,17.2,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-4.7 blog: glm-4.6=17.2.
glm-4.6,aime_2025,93.9,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-4.7 blog: glm-4.6=93.9.
glm-4.6,livecodebench,70.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2 model card: glm-4.6=70.
glm-4.6,mmlu_pro,83.2,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-4.7 blog: glm-4.6=83.2.
glm-4.6,terminal_bench,24.5,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-4.7 blog: glm-4.6=24.5.
glm-4.6,humaneval,82.0,https://llm-stats.com/models/glm-4.7,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
glm-4.6,ifeval,82.0,https://llm-stats.com/models/glm-4.7,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
glm-4.6,mmlu,85.0,https://llm-stats.com/models/glm-4.7,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
glm-4.7,aime_2025,95.7,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-4.7 blog: glm-4.7=95.7.
glm-4.7,gpqa_diamond,85.7,https://z.ai/blog/glm-5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5 blog: glm-4.7=85.7.
glm-4.7,hle,24.8,https://z.ai/blog/glm-5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5 blog: glm-4.7=24.8.
glm-4.7,swe_bench_verified,73.8,https://z.ai/blog/glm-5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,GLM-5 blog: glm-4.7=73.8.
glm-4.7,livecodebench,84.9,https://vertu.com/lifestyle/glm-4-7-released/,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
glm-4.7,chatbot_arena_elo,1443.0,https://lmarena.ai/leaderboard/text,{},True,leaderboard,verified,lmarena.ai text arena (fetched 2026-04-29). Arena model name: glm-4.7 (rank 48). ELO updates continuously; score reflects latest available.
glm-4.7,browsecomp,52.0,https://z.ai/blog/glm-5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5 blog: glm-4.7=52.0.
glm-4.7,frontiermath,2.4,https://epoch.ai/benchmarks/frontiermath,"{""effort"":""default"",""harness"":""Together API"",""mode"":""thinking"",""tools"":""none""}",True,leaderboard,verified,"epoch.ai FrontierMath-2025-02-28-Private: zai-org/GLM-4.7 (GLM-4.7 Together) = 2.4%. Canonical thinking mode matches. BP had 20.0 (unverified; likely sourced from wrong model — canonical notes reference GLM-5 blog, but model is GLM-4.7)."
glm-4.7,mmlu_pro,84.3,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-4.7 blog: glm-4.7=84.3.
glm-4.7,terminal_bench,41.0,https://z.ai/blog/glm-5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5 blog: glm-4.7=41.0.
glm-4.7,livebench,57.3,https://livebench.ai/,"{""effort"":""default"",""mode"":""thinking"",""tools"":""n/a""}",True,leaderboard,verified,livebench.ai table_2026_01_08 (latest as of 2026-04-28): glm-4.7 avg=57.3. Canonical thinking mode matches. BP had 58.1.
glm-4.7,simpleqa,32.0,https://llm-stats.com/models/glm-4.7,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
glm-5.1,aime_2026,95.3,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5.1 blog: glm-5.1=95.3.
glm-5.1,gpqa_diamond,86.2,https://z.ai/blog/glm-5.2,"{""context"":""default"",""effort"":""official reported"",""harness"":""official reasoning evaluation"",""judge"":""GPT-5.5 medium for AIME/HMMT/IMO; benchmark-specified otherwise"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none"",""top_p"":""0.95""}",True,official_blog,verified,GLM-5.2 official release source: GLM-5.1 / GPQA-Diamond = 86.2.
glm-5.1,swe_bench_pro,58.4,https://z.ai/blog/glm-5.2,"{""context"":""400000"",""effort"":""official reported"",""harness"":""OpenHands with tailored instruction prompt"",""judge"":""official tests"",""max_output_tokens"":""32000"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic code execution"",""top_p"":""1.0""}",True,official_blog,verified,GLM-5.2 official release source: GLM-5.1 / SWE-bench Pro = 58.4.
glm-5.1,terminal_bench,63.5,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): glm-5.1 (per their own canonical config) = 63.5.
glm-5.1,browsecomp,79.3,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): glm-5.1 (per their own canonical config) = 79.3.
doubao-seed-2.0-pro,aime_2025,98.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=98.3.
doubao-seed-2.0-pro,gpqa_diamond,88.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=88.9.
doubao-seed-2.0-pro,swe_bench_verified,76.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=76.5.
doubao-seed-2.0-pro,browsecomp,77.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=77.3.
doubao-seed-2.0-pro,video_mmmu,86.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=86.9.
doubao-seed-2.0-pro,mmmu,85.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=85.4.
doubao-seed-2.0-pro,mathvision,88.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=88.8.
doubao-seed-2.0-pro,codeforces_rating,3020.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=3020.
doubao-seed-2.0-pro,hle,32.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=32.4.
doubao-seed-2.0-pro,livecodebench,88.0,https://www.digitalapplied.com/blog/bytedance-seed-2-doubao-ai-model-benchmarks-guide,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
doubao-seed-2.0-pro,mmlu_pro,87.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=87.0.
doubao-seed-2.0-pro,aime_2024,98.3,https://www.digitalapplied.com/blog/bytedance-seed-2-doubao-ai-model-benchmarks-guide,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
doubao-seed-2.0-pro,humaneval,92.0,https://www.digitalapplied.com/blog/bytedance-seed-2-doubao-ai-model-benchmarks-guide,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
doubao-seed-2.0-pro,ifeval,88.5,https://www.digitalapplied.com/blog/bytedance-seed-2-doubao-ai-model-benchmarks-guide,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
doubao-seed-2.0-pro,mmlu,90.0,https://www.digitalapplied.com/blog/bytedance-seed-2-doubao-ai-model-benchmarks-guide,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
doubao-seed-2.0-pro,terminal_bench,55.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=55.8.
mistral-small-3.1,mmlu,80.62,https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Instruct-2503,"{""harness"":""mistral-eval"",""judge"":""rule-based"",""prompt_style"":""5-shot CoT"",""sampling"":""pass@1"",""temperature"":""0.0""}",True,model_card,verified,HF model card Table 1: MMLU = 80.62%.
mistral-small-3.1,humaneval,88.41,https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Instruct-2503,"{""harness"":""mistral-eval"",""judge"":""rule-based"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.0""}",True,model_card,verified,HF model card Table 1: HumanEval = 88.41%. BP previously had 88.99 (secondary blog).
mistral-small-3.1,ifeval,82.75,,{},False,,dropped,Mistral does NOT report IFEval in HF MC or blog. BP value 82.75 came from venturebeat secondary blog.
mistral-small-3.1,mmlu_pro,66.76,https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Instruct-2503,"{""harness"":""mistral-eval"",""judge"":""rule-based"",""prompt_style"":""5-shot CoT"",""sampling"":""pass@1"",""temperature"":""0.0""}",True,model_card,verified,HF model card Table 1: MMLU Pro (5-shot CoT) = 66.76%.
mistral-small-3.1,gpqa_diamond,45.96,https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Instruct-2503,"{""harness"":""mistral-eval"",""judge"":""rule-based"",""prompt_style"":""5-shot CoT"",""sampling"":""pass@1"",""temperature"":""0.0""}",True,model_card,verified,HF model card Table 1: GPQA Diamond (5-shot CoT) = 45.96%. BP previously had 39.0 (secondary blog).
mistral-small-3.1,livecodebench,30.0,,{},False,,dropped,Mistral does NOT report LiveCodeBench in HF MC or blog. BP value 30.0 came from venturebeat secondary blog.
mistral-small-3.1,math_500,81.0,,{},False,,dropped,Mistral does NOT report MATH-500. HF Table 1 has only 'MATH' = 69.30%. BP value 81.0 came from venturebeat secondary blog (unverifiable). Use the new 'math' cell instead.
mistral-small-3.1,aime_2024,22.0,,{},False,,dropped,Mistral does NOT report AIME in HF MC or blog. BP value 22 came from analyticsvidhya secondary blog.
mistral-small-3.1,simpleqa,10.43,https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Instruct-2503,"{""harness"":""mistral-eval"",""judge"":""rule-based"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.0""}",True,model_card,verified,HF model card Table 1: SimpleQA TotalAcc = 10.43%. BP previously had 18 (secondary blog).
mistral-small-3.1,swe_bench_verified,28.0,,{},False,,dropped,Mistral does NOT report SWE-bench Verified in HF MC or blog. BP value 28 came from analyticsvidhya secondary blog.
mistral-medium-3,humaneval,92.1,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",True,official_blog,verified,Mistral Medium 3 blog table: 92.1.
mistral-medium-3,arena_hard,97.1,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",True,official_blog,verified,Mistral Medium 3 blog table: 97.1.
mistral-medium-3,math_500,91.0,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",True,official_blog,verified,Mistral Medium 3 blog table: 91.0.
mistral-medium-3,gpqa_diamond,57.1,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""5-shot CoT"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",True,official_blog,verified,Mistral Medium 3 blog table: 57.1.
mistral-medium-3,ifeval,89.4,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",True,official_blog,verified,Mistral Medium 3 blog table: 89.4.
mistral-medium-3,mmlu,87.0,https://apidog.com/blog/mistral-medium-3/,{},,,dropped,DROPPED 2026-04-29T02:20: Not in Mistral Medium 3 official blog table; previous ref was secondary/third-party.
mistral-medium-3,livecodebench,42.0,https://artificialanalysis.ai/models/mistral-medium-3-1,{},,,dropped,DROPPED 2026-04-29T02:20: Not in Mistral Medium 3 official blog table; previous ref was secondary/third-party.
mistral-medium-3,mmlu_pro,77.2,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""5-shot CoT"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",True,official_blog,verified,Mistral Medium 3 blog table: 77.2.
mistral-medium-3,simpleqa,25.0,https://artificialanalysis.ai/models/mistral-medium-3-1,{},,,dropped,DROPPED 2026-04-29T02:20: Not in Mistral Medium 3 official blog table; previous ref was secondary/third-party.
mistral-medium-3,swe_bench_verified,32.0,https://artificialanalysis.ai/models/mistral-medium-3-1,{},,,dropped,DROPPED 2026-04-29T02:20: Not in Mistral Medium 3 official blog table; previous ref was secondary/third-party.
mistral-large-3,mmlu_pro,80.9,https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/resolve/a11f36bebf709121056b1dbcc943d1c6afbe494d/images/image3.png?download=true,"{""context"":""source does not state"",""effort"":""none"",""harness"":""Mistral official model-card evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official/default or source does not state"",""sampling"":""pass@1; repeat count undisclosed"",""temperature"":""source does not state"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly: '80.9'. Research observation: small-image3.png:2:4:score:reported.
mistral-large-3,math_500,93.6,https://intuitionlabs.ai/articles/mistral-large-3-moe-llm-explained,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
mistral-large-3,aime_2025,85.0,https://www.analyticsvidhya.com/blog/2025/12/mistral-large-3/,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
mistral-large-3,swe_bench_verified,68.0,https://llm-stats.com/benchmarks/swe-bench-verified,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
mistral-large-3,gpqa_diamond,64.1,https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/resolve/a11f36bebf709121056b1dbcc943d1c6afbe494d/images/image3.png?download=true,"{""context"":""source does not state"",""effort"":""none"",""harness"":""Mistral official model-card evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official/default or source does not state"",""sampling"":""pass@1; repeat count undisclosed"",""temperature"":""source does not state"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly: '64.1'. Research observation: small-image3.png:1:4:score:reported.
mistral-large-3,humaneval,92.0,https://medium.com/@leucopsis/mistral-large-3-2512-review-7788c779a5e4,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
mistral-large-3,ifeval,86.0,https://intuitionlabs.ai/articles/mistral-large-3-moe-llm-explained,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
mistral-large-3,livecodebench,66.0,https://medium.com/@leucopsis/mistral-large-3-2512-review-7788c779a5e4,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
mistral-large-3,mmlu,85.5,https://medium.com/@leucopsis/mistral-large-3-2512-review-7788c779a5e4,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
mistral-large-3,aime_2024,53.3,https://medium.com/@leucopsis/mistral-large-3-2512-review-7788c779a5e4,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
mistral-large-3,chatbot_arena_elo,1418.0,https://medium.com/@leucopsis/mistral-large-3-2512-review-7788c779a5e4,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
mistral-large-3,codeforces_rating,1550.0,https://medium.com/@leucopsis/mistral-large-3-2512-review-7788c779a5e4,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
mistral-large-3,simpleqa,30.0,https://medium.com/@leucopsis/mistral-large-3-2512-review-7788c779a5e4,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
codestral-25.01,humaneval,86.6,https://blog.getbind.co/2025/01/15/mistral-codestral-25-01/,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
codestral-25.01,livecodebench,37.9,https://blog.getbind.co/2025/01/15/mistral-codestral-25-01/,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
codestral-25.01,math_500,74.0,https://blog.getbind.co/2025/01/15/mistral-codestral-25-01/,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
codestral-25.01,swe_bench_verified,35.0,https://llm-stats.com/benchmarks,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
codestral-25.01,codeforces_rating,1480.0,https://llm-stats.com/benchmarks,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
devstral-2,swe_bench_verified,72.2,https://mistral.ai/news/devstral-2-vibe-cli,{},,official_blog,verified,
devstral-2,humaneval,88.0,https://mistral.ai/news/devstral-2-vibe-cli,{},,,dropped,
devstral-2,livecodebench,60.0,https://mistral.ai/news/devstral-2-vibe-cli,{},,,dropped,
devstral-2,math_500,72.0,https://llm-stats.com/benchmarks,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
devstral-2,gpqa_diamond,45.0,https://llm-stats.com/benchmarks,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
phi-4,mmlu,84.8,https://arxiv.org/abs/2412.08905,"{""effort"":""n/a"",""harness"":""simple-evals"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.5"",""tools"":""none""}",True,tech_report,verified,"Phi-4 paper Table 1 (phi-4 column): mmlu=84.8 (re-sourced from https://arxiv.org/abs/2412.08905, matches BP value)"
phi-4,gpqa_diamond,56.1,https://arxiv.org/abs/2412.08905,"{""effort"":""n/a"",""harness"":""simple-evals"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.5"",""tools"":""none""}",True,tech_report,verified,"Phi-4 paper Table 1 (phi-4 column): gpqa_diamond=56.1 (re-sourced from https://arxiv.org/abs/2412.08905, matches BP value)"
phi-4,humaneval,82.6,https://arxiv.org/abs/2412.08905,"{""effort"":""n/a"",""harness"":""simple-evals"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.5"",""tools"":""none""}",True,tech_report,verified,"Phi-4 paper Table 1 (phi-4 column): humaneval=82.6 (re-sourced from https://arxiv.org/abs/2412.08905, matches BP value)"
phi-4,arena_hard,75.4,https://arxiv.org/abs/2412.08905,"{""effort"":""n/a"",""harness"":""simple-evals"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.5"",""tools"":""none""}",True,tech_report,verified,"Phi-4 paper Table 1 (phi-4 column): arena_hard=75.4 (re-sourced from https://arxiv.org/abs/2412.08905, matches BP value)"
phi-4,math_500,80.4,https://arxiv.org/abs/2412.08905,"{""effort"":""n/a"",""harness"":""simple-evals"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.5"",""tools"":""none""}",True,tech_report,verified,"Phi-4 paper Table 1 (phi-4 column): math_500=80.4 (re-sourced from https://arxiv.org/abs/2412.08905, matches BP value)"
phi-4,ifeval,63.0,https://arxiv.org/abs/2412.08905,"{""effort"":""n/a"",""harness"":""simple-evals"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.5"",""tools"":""none""}",True,tech_report,verified,Phi-4 paper Table 1 (phi-4 column): ifeval=63.0. [R5d: prior unverified value 79.3 from https://arxiv.org/html/2412.08905v1 deleted.]
phi-4,aime_2024,18.0,https://huggingface.co/microsoft/phi-4,{},,,dropped,
phi-4,bigcodebench,45.5,https://bigcode-bench.github.io/,"""BigCodeBench Complete Instruct pass@1""",True,leaderboard,verified,
phi-4,livecodebench,38.5,https://huggingface.co/microsoft/phi-4,{},,,dropped,
phi-4,mmlu_pro,70.4,https://arxiv.org/abs/2412.08905,"{""effort"":""n/a"",""harness"":""simple-evals"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.5"",""tools"":""none""}",True,tech_report,verified,Phi-4 paper Table 1 (phi-4 column): mmlu_pro=70.4. [R5d: prior unverified value 68 from https://huggingface.co/microsoft/phi-4 deleted.]
phi-4,simpleqa,3.0,https://arxiv.org/abs/2412.08905,"{""effort"":""n/a"",""harness"":""simple-evals"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.5"",""tools"":""none""}",True,tech_report,verified,Phi-4 paper Table 1 (phi-4 column): simpleqa=3.0. [R5d: prior unverified value 15 from https://huggingface.co/microsoft/phi-4 deleted.]
phi-4-reasoning,aime_2025,63.1,https://arxiv.org/abs/2504.21318,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.8"",""tools"":""none""}",True,tech_report,verified,Phi-4-reasoning paper Table 1 (phi-4-reasoning column): aime_2025=63.1. [R5d: prior unverified value 71.4 from https://www.analyticsvidhya.com/blog/2025/05/phi-4-reasoning-models/ deleted.]
phi-4-reasoning,gpqa_diamond,67.1,https://arxiv.org/abs/2504.21318,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.8"",""tools"":""none""}",True,tech_report,verified,Phi-4-reasoning paper Table 1 (phi-4-reasoning column): gpqa_diamond=67.1. [R5d: prior unverified value 63.4 from https://ashishchadha11944.medium.com/microsofts-phi-4-reasoning-models/ deleted.]
phi-4-reasoning,livecodebench,53.8,https://arxiv.org/abs/2504.21318,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.8"",""tools"":""none""}",True,tech_report,verified,"Phi-4-reasoning paper Table 1 (phi-4-reasoning column): livecodebench=53.8 (re-sourced from https://arxiv.org/abs/2504.21318, matches BP value)"
phi-4-reasoning,math_500,95.0,https://www.microsoft.com/en-us/research/articles/phi-reasoning/,{},,,dropped,
phi-4-reasoning,aime_2024,74.6,https://arxiv.org/abs/2504.21318,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.8"",""tools"":""none""}",True,tech_report,verified,Phi-4-reasoning paper Table 1 (phi-4-reasoning column): aime_2024=74.6. [R5d: prior unverified value 70.0 from https://www.microsoft.com/en-us/research/wp-content/uploads/2025/04/phi_4_reasoning.pdf deleted.]
phi-4-reasoning,mmlu_pro,74.3,https://arxiv.org/abs/2504.21318,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.8"",""tools"":""none""}",True,tech_report,verified,Phi-4-reasoning paper Table 2 (phi-4-reasoning column): mmlu_pro=74.3. [R5d: prior unverified value 72.0 from https://huggingface.co/microsoft/Phi-4-reasoning deleted.]
phi-4-reasoning,humaneval,85.0,https://huggingface.co/microsoft/Phi-4-reasoning,{},,,dropped,
phi-4-reasoning,codeforces_rating,1736.0,https://arxiv.org/abs/2504.21318,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.8"",""tools"":""none""}",True,tech_report,verified,Phi-4-reasoning paper Table 1 (phi-4-reasoning column): codeforces_rating=1736. [R5d: prior unverified value 1500 from https://huggingface.co/microsoft/Phi-4-reasoning deleted.]
phi-4-reasoning,hmmt_feb_2025,43.8,https://arxiv.org/abs/2504.21318,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.8"",""tools"":""none""}",True,tech_report,verified,"Phi-4-reasoning paper Table 1 (phi-4-reasoning column): hmmt_2025=43.8 (re-sourced from https://arxiv.org/abs/2504.21318, matches BP value)"
phi-4-reasoning,ifeval,83.4,https://arxiv.org/abs/2504.21318,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.8"",""tools"":""none""}",True,tech_report,verified,"Phi-4-reasoning paper Table 2 (phi-4-reasoning column): ifeval=83.4 (re-sourced from https://arxiv.org/abs/2504.21318, matches BP value)"
phi-4-reasoning,arena_hard,73.3,https://arxiv.org/abs/2504.21318,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.8"",""tools"":""none""}",True,tech_report,verified,"Phi-4-reasoning paper Table 2 (phi-4-reasoning column): arena_hard=73.3 (re-sourced from https://arxiv.org/abs/2504.21318, matches BP value)"
phi-4-reasoning-plus,aime_2025,78.0,https://arxiv.org/abs/2504.21318,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.8"",""tools"":""none""}",True,tech_report,verified,Phi-4-reasoning paper Table 1 (phi-4-reasoning-plus column): aime_2025=78.0. [R5d: prior unverified value 77.7 from https://huggingface.co/microsoft/Phi-4-reasoning-plus deleted.]
phi-4-reasoning-plus,gpqa_diamond,69.3,https://arxiv.org/abs/2504.21318,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.8"",""tools"":""none""}",True,tech_report,verified,"Phi-4-reasoning paper Table 1 (phi-4-reasoning-plus column): gpqa_diamond=69.3 (re-sourced from https://arxiv.org/abs/2504.21318, matches BP value)"
phi-4-reasoning-plus,livecodebench,53.1,https://arxiv.org/abs/2504.21318,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.8"",""tools"":""none""}",True,tech_report,verified,Phi-4-reasoning paper Table 1 (phi-4-reasoning-plus column): livecodebench=53.1. [R5d: prior unverified value 68.8 from https://ashishchadha11944.medium.com/microsofts-phi-4-reasoning-models/ deleted.]
phi-4-reasoning-plus,math_500,96.4,https://huggingface.co/microsoft/Phi-4-reasoning-plus,{},,,dropped,
phi-4-reasoning-plus,humaneval,88.0,https://huggingface.co/microsoft/Phi-4-reasoning-plus,{},,,dropped,
phi-4-reasoning-plus,aime_2024,81.3,https://arxiv.org/abs/2504.21318,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.8"",""tools"":""none""}",True,tech_report,verified,Phi-4-reasoning paper Table 1 (phi-4-reasoning-plus column): aime_2024=81.3. [R5d: prior unverified value 70 from https://www.microsoft.com/en-us/research/wp-content/uploads/2025/04/phi_4_reasoning.pdf deleted.]
phi-4-reasoning-plus,arena_hard,79.0,https://arxiv.org/abs/2504.21318,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.8"",""tools"":""none""}",True,tech_report,verified,"Phi-4-reasoning paper Table 2 (phi-4-reasoning-plus column): arena_hard=79.0 (re-sourced from https://arxiv.org/abs/2504.21318, matches BP value)"
phi-4-reasoning-plus,codeforces_rating,1723.0,https://arxiv.org/abs/2504.21318,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.8"",""tools"":""none""}",True,tech_report,verified,Phi-4-reasoning paper Table 1 (phi-4-reasoning-plus column): codeforces_rating=1723. [R5d: prior unverified value 1500 from https://huggingface.co/microsoft/Phi-4-reasoning deleted.]
phi-4-reasoning-plus,hmmt_feb_2025,53.6,https://arxiv.org/abs/2504.21318,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.8"",""tools"":""none""}",True,tech_report,verified,"Phi-4-reasoning paper Table 1 (phi-4-reasoning-plus column): hmmt_2025=53.6 (re-sourced from https://arxiv.org/abs/2504.21318, matches BP value)"
phi-4-reasoning-plus,ifeval,84.9,https://arxiv.org/abs/2504.21318,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.8"",""tools"":""none""}",True,tech_report,verified,Phi-4-reasoning paper Table 2 (phi-4-reasoning-plus column): ifeval=84.9. [R5d: prior unverified value 82 from https://huggingface.co/microsoft/Phi-4-reasoning-plus deleted.]
phi-4-reasoning-plus,mmlu_pro,76.0,https://arxiv.org/abs/2504.21318,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.8"",""tools"":""none""}",True,tech_report,verified,Phi-4-reasoning paper Table 2 (phi-4-reasoning-plus column): mmlu_pro=76.0. [R5d: prior unverified value 72 from https://huggingface.co/microsoft/Phi-4-reasoning deleted.]
phi-4-reasoning-plus,swe_bench_verified,32.0,https://huggingface.co/microsoft/Phi-4-reasoning-plus,{},,,dropped,
nemotron-ultra-253b,gpqa_diamond,76.01,https://huggingface.co/nvidia/Llama-3_1-Nemotron-Ultra-253B-v1,"{""context"":""32k"",""effort"":""default"",""harness"":""NVIDIA eval pipeline (transformers/vLLM)"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""custom prompt templates per benchmark"",""sampling"":""pass@1 (avg of up to 16 benchmark runs)"",""temperature"":""0.6"",""tools"":""none""}",True,model_card,verified,"HF model card ""GPQA"" Reasoning On=76.01. NOTE: model card says ""GPQA"" not ""GPQA Diamond"" — likely same but verify (see review file). Prompt template = ABCD format consistent with Diamond."
nemotron-ultra-253b,aime_2025,72.5,https://huggingface.co/nvidia/Llama-3_1-Nemotron-Ultra-253B-v1,"{""context"":""32k"",""effort"":""default"",""harness"":""NVIDIA eval pipeline (transformers/vLLM)"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""custom prompt templates per benchmark"",""sampling"":""pass@1 (avg of up to 16 benchmark runs)"",""temperature"":""0.6"",""tools"":""none""}",True,model_card,verified,"HF model card ""AIME25"" Reasoning On=72.50. Clear match."
nemotron-ultra-253b,livecodebench,66.31,https://huggingface.co/nvidia/Llama-3_1-Nemotron-Ultra-253B-v1,"{""context"":""32k"",""effort"":""default"",""harness"":""NVIDIA eval pipeline (transformers/vLLM)"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""custom prompt templates per benchmark"",""sampling"":""pass@1 (avg of up to 16 benchmark runs)"",""temperature"":""0.6"",""tools"":""none""}",True,model_card,verified,"HF model card ""LiveCodeBench (20240801-20250201)"" Reasoning On=66.31. Date range noted in reported_setting."
nemotron-ultra-253b,ifeval,89.45,https://huggingface.co/nvidia/Llama-3_1-Nemotron-Ultra-253B-v1,"{""context"":""32k"",""effort"":""default"",""harness"":""NVIDIA eval pipeline (transformers/vLLM)"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""custom prompt templates per benchmark"",""sampling"":""pass@1 (avg of up to 16 benchmark runs)"",""temperature"":""0.6"",""tools"":""none""}",True,model_card,verified,"HF model card ""IFEval"" Strict:Instruction Reasoning On=89.45. Clear match."
nemotron-ultra-253b,math_500,97.0,https://huggingface.co/nvidia/Llama-3_1-Nemotron-Ultra-253B-v1,"{""context"":""32k"",""effort"":""default"",""harness"":""NVIDIA eval pipeline (transformers/vLLM)"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""custom prompt templates per benchmark"",""sampling"":""pass@1 (avg of up to 16 benchmark runs)"",""temperature"":""0.6"",""tools"":""none""}",True,model_card,verified,"HF model card ""MATH500"" Reasoning On=97.00. Matches exactly. Previous ref was 404 URL; updated to HF model card."
nemotron-ultra-253b,arena_hard,92.7,https://medium.com/towards-agi/nvidia-llama-nemotron-outshines-llama-4/,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
nemotron-ultra-253b,humaneval,92.0,https://developer.nvidia.com/blog/nvidia-llama-nemotron-ultra-open-model-delivers-groundbreaking-reasoning-accuracy/,{},,,dropped,"DROPPED (R5g-ghost): original ref URL was dead 404. Correct URL is https://developer.nvidia.com/blog/nvidia-llama-nemotron-ultra-open-model-delivers-groundbreaking-reasoning-accuracy/. That blog covers GPQA Diamond (76%), LiveCodeBench, AIME only — no humaneval data for this model."
nemotron-ultra-253b,mmlu,87.0,https://developer.nvidia.com/blog/nvidia-llama-nemotron-ultra-open-model-delivers-groundbreaking-reasoning-accuracy/,{},,,dropped,"DROPPED (R5g-ghost): original ref URL was dead 404. Correct URL is https://developer.nvidia.com/blog/nvidia-llama-nemotron-ultra-open-model-delivers-groundbreaking-reasoning-accuracy/. That blog covers GPQA Diamond (76%), LiveCodeBench, AIME only — no mmlu data for this model."
nemotron-ultra-253b,mmlu_pro,76.0,https://developer.nvidia.com/blog/nvidia-llama-nemotron-ultra-open-model-delivers-groundbreaking-reasoning-accuracy/,{},,,dropped,"DROPPED (R5g-ghost): original ref URL was dead 404. Correct URL is https://developer.nvidia.com/blog/nvidia-llama-nemotron-ultra-open-model-delivers-groundbreaking-reasoning-accuracy/. That blog covers GPQA Diamond (76%), LiveCodeBench, AIME only — no mmlu_pro data for this model."
nemotron-ultra-253b,codeforces_rating,1750.0,https://artificialanalysis.ai/models/llama-3-1-nemotron-ultra-253b-v1-reasoning,{},,,needs_review,Reference URL is AA model page (llama-3-1-nemotron-ultra-253b-v1-reasoning). AA does not publish codeforces_rating data; field is not present in AA model data. Cannot verify 1750 from AA source. Original source may be NVIDIA tech report.
nemotron-ultra-253b,hle,15.0,https://artificialanalysis.ai/models/llama-3-1-nemotron-ultra-253b-v1-reasoning,{},,,needs_review,"Reference URL is AA model page (llama-3-1-nemotron-ultra-253b-v1-reasoning). Current AA model page shows hle=0.081→8.1%, but BP has 15. Major mismatch (6.9pp). Possible stale data or wrong model variant. Current AA value: 8.1%."
nemotron-ultra-253b,simpleqa,35.0,https://artificialanalysis.ai/models/llama-3-1-nemotron-ultra-253b-v1-reasoning,{},,,needs_review,Reference URL is AA model page (llama-3-1-nemotron-ultra-253b-v1-reasoning). AA does not publish SimpleQA scores; field is not present in AA model data for any model. Cannot verify from AA source. Original source needs identification.
nemotron-ultra-253b,swe_bench_verified,48.0,https://artificialanalysis.ai/models/llama-3-1-nemotron-ultra-253b-v1-reasoning,{},,,needs_review,Reference URL is AA model page (llama-3-1-nemotron-ultra-253b-v1-reasoning). AA does not publish swe_bench_verified scores; field is not present in AA model data. Cannot verify 48 from AA source.
amazon-nova-pro,mmlu,85.9,https://cdn.amazon.science/e5/e6/ccc5378c42dca467d1abe1628ec9/amazon-nova-premier-technical-report-and-model-card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""zeroshot-cot"",""sampling"":""pass@1"",""temperature"":0.0,""tools"":""none""}",True,tech_report,verified,
amazon-nova-pro,gpqa_diamond,50.0,https://cdn.amazon.science/e5/e6/ccc5378c42dca467d1abe1628ec9/amazon-nova-premier-technical-report-and-model-card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""zeroshot-cot"",""sampling"":""pass@1"",""temperature"":0.0,""tools"":""none""}",True,tech_report,verified,
amazon-nova-pro,humaneval,89.0,https://arxiv.org/abs/2506.12103,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""zeroshot-cot"",""sampling"":""pass@1"",""temperature"":0.0,""tools"":""none""}",True,tech_report,verified,From Nova family tech report (arxiv 2506.12103); not in Premier announcement table.
amazon-nova-pro,ifeval,92.1,https://cdn.amazon.science/e5/e6/ccc5378c42dca467d1abe1628ec9/amazon-nova-premier-technical-report-and-model-card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""zeroshot-cot"",""sampling"":""pass@1"",""temperature"":0.0,""tools"":""none""}",True,tech_report,verified,
amazon-nova-pro,math_500,76.6,https://cdn.amazon.science/e5/e6/ccc5378c42dca467d1abe1628ec9/amazon-nova-premier-technical-report-and-model-card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""zeroshot-cot"",""sampling"":""pass@1"",""temperature"":0.0,""tools"":""none""}",True,tech_report,verified,
amazon-nova-premier,mmlu,87.4,https://cdn.amazon.science/e5/e6/ccc5378c42dca467d1abe1628ec9/amazon-nova-premier-technical-report-and-model-card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""zeroshot-cot"",""sampling"":""pass@1"",""temperature"":0.0,""tools"":""none""}",True,tech_report,verified,
amazon-nova-premier,math_500,82.0,https://cdn.amazon.science/e5/e6/ccc5378c42dca467d1abe1628ec9/amazon-nova-premier-technical-report-and-model-card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""zeroshot-cot"",""sampling"":""pass@1"",""temperature"":0.0,""tools"":""none""}",True,tech_report,verified,
amazon-nova-premier,swe_bench_verified,42.4,https://cdn.amazon.science/e5/e6/ccc5378c42dca467d1abe1628ec9/amazon-nova-premier-technical-report-and-model-card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""zeroshot-cot"",""sampling"":""pass@1"",""temperature"":0.0,""tools"":""none""}",True,tech_report,verified,
amazon-nova-premier,gpqa_diamond,57.1,https://cdn.amazon.science/e5/e6/ccc5378c42dca467d1abe1628ec9/amazon-nova-premier-technical-report-and-model-card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""zeroshot-cot"",""sampling"":""pass@1"",""temperature"":0.0,""tools"":""none""}",True,tech_report,verified,
amazon-nova-premier,humaneval,80.0,https://aws.amazon.com/blogs/aws/amazon-nova-premier/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""zeroshot-cot"",""sampling"":""pass@1"",""temperature"":0.0,""tools"":""none""}",True,official_blog,dropped,Cited in AWS announcement blog; no detailed methodology disclosed. Pending decision.
amazon-nova-premier,ifeval,91.5,https://cdn.amazon.science/e5/e6/ccc5378c42dca467d1abe1628ec9/amazon-nova-premier-technical-report-and-model-card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""zeroshot-cot"",""sampling"":""pass@1"",""temperature"":0.0,""tools"":""none""}",True,tech_report,verified,
amazon-nova-premier,mmmu,68.0,https://cdn.amazon.science/e5/e6/ccc5378c42dca467d1abe1628ec9/amazon-nova-premier-technical-report-and-model-card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""zeroshot-cot"",""sampling"":""pass@1"",""temperature"":0.0,""tools"":""none""}",True,tech_report,verified,
command-a,mmlu,85.5,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,"Command A paper: mmlu=85.5 (matches BP, re-sourced)"
command-a,humaneval,82.9,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 82.9. Third-party self-test by Mistral (their internal eval pipeline).
command-a,mmlu_pro,68.9,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""5-shot CoT"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 68.9. Third-party self-test by Mistral (their internal eval pipeline).
command-a,ifeval,89.7,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 89.7. Third-party self-test by Mistral (their internal eval pipeline).
command-a,arena_hard,95.1,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,dropped,Mistral Medium 3 blog table: 95.1. Third-party self-test by Mistral (their internal eval pipeline). | DROPPED: BP cell URL = Cohere paper but paper main tables do not report standard ArenaHard. Paper has mArenaHard (multilingual) and Auto Arena Hard appendix figure but not the cell value 72.0. Ghost per R5g(a).
command-a,gpqa_diamond,46.5,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""5-shot CoT"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,dropped,"Mistral Medium 3 blog table: 46.5. Third-party self-test by Mistral (their internal eval pipeline). | DROPPED: BP cell URL = Cohere paper but paper Table 1/3 reports GPQA full set 50.8, NOT GPQA Diamond. Wrong-benchmark scrape (ghost) per R5g(b). Use new gpqa cell at 50.8."
command-a,livecodebench,26.9,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Command A paper: livecodebench=26.9. [R5d: prior unverified value 28.0 from https://cohere.com/research/papers/command-a-technical-report.pdf deleted.]
command-a,math_500,82.0,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 82.0. Third-party self-test by Mistral (their internal eval pipeline).
command-a,aime_2024,30.0,https://www.vellum.ai/llm-leaderboard,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
command-a,bigcodebench,45.4,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Command A paper: bigcodebench=45.4. [R5d: prior unverified value 33.8 from https://bigcode-bench.github.io/ deleted.]
command-a,simpleqa,32.0,https://www.vellum.ai/llm-leaderboard,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
command-a,swe_bench_verified,26.8,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Command A paper: swe_bench_verified=26.8. [R5d: prior unverified value 38 from https://www.vellum.ai/llm-leaderboard deleted.]
exaone-4.0-32b,mmlu_pro,81.8,https://arxiv.org/html/2507.11407v1,"{""effort"":""default"",""harness"":""LG AI Research internal"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.6, top_p=0.95, presence_penalty=1.5"",""tools"":""none""}",True,tech_report,verified,Reasoning mode Table 3: MMLU-Pro=81.8 matches BP.
exaone-4.0-32b,gpqa_diamond,75.4,https://arxiv.org/html/2507.11407v1,"{""effort"":""default"",""harness"":""LG AI Research internal"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg@8"",""temperature"":""0.6, top_p=0.95, presence_penalty=1.5"",""tools"":""none""}",True,tech_report,verified,Reasoning mode Table 3: GPQA-Diamond=75.4 matches BP.
exaone-4.0-32b,aime_2025,85.3,https://arxiv.org/html/2507.11407v1,"{""effort"":""default"",""harness"":""LG AI Research internal"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg@32"",""temperature"":""0.6, top_p=0.95, presence_penalty=1.5"",""tools"":""none""}",True,tech_report,verified,"Reasoning mode Table 3: AIME 2025=85.3 matches BP. (avg@32, 64K budget)"
exaone-4.0-32b,livecodebench,72.6,https://arxiv.org/html/2507.11407v1,{},False,tech_report,dropped,DROPPED (R5g-b wrong-variant): Source explicitly labels this LiveCodeBench v5=72.6. BP has separate livecodebench_v5 and livecodebench_v6 IDs. Moved to livecodebench_v5.
exaone-4.0-32b,hmmt_feb_2025,72.9,https://arxiv.org/html/2507.11407v1,"{""effort"":""default"",""harness"":""LG AI Research internal"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg@32"",""temperature"":""0.6, top_p=0.95, presence_penalty=1.5"",""tools"":""none""}",True,tech_report,verified,Reasoning mode Table 3: HMMT Feb 2025=72.9 matches BP. (avg@32)
exaone-4.0-32b,math_500,96.4,https://arxiv.org/html/2507.11407v1,{},False,tech_report,dropped,DROPPED (R5g-a ghost): Source only has MATH500(ES)=95.8 (Reasoning). English MATH-500 not reported. 96.4 not found in source.
exaone-4.0-32b,ifeval,83.7,https://arxiv.org/html/2507.11407v1,"{""effort"":""default"",""harness"":""LG AI Research internal"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (greedy for Non-Reasoning, sampled for Reasoning avg)"",""temperature"":""0.6, top_p=0.95, presence_penalty=1.5"",""tools"":""none""}",True,tech_report,verified,Reasoning mode Table 3: IFEval=83.7. BP had 88.0 which is not present anywhere in source → R5g-b overwrite.
exaone-4.0-32b,humaneval,90.2,https://arxiv.org/html/2507.11407v1,{},False,tech_report,dropped,DROPPED (R5g-a ghost): HumanEval not evaluated in EXAONE 4.0 tech report (2507.11407). Source has no HumanEval table.
exaone-4.0-32b,arena_hard,80.0,https://arxiv.org/html/2507.11407v1,{},False,tech_report,dropped,DROPPED (R5g-a ghost): Arena-Hard not evaluated in EXAONE 4.0 tech report. Source has no Arena Hard table.
exaone-4.0-32b,mmlu,84.0,https://arxiv.org/html/2507.11407v1,{},False,tech_report,dropped,"DROPPED (R5g-a ghost): Source uses MMLU-Redux (92.3) and MMLU-Pro (81.8), not plain MMLU. 84.0 not found in source."
exaone-4.0-32b,simpleqa,30.0,https://arxiv.org/html/2507.11407v1,{},False,tech_report,dropped,DROPPED (R5g-a ghost): SimpleQA not evaluated in EXAONE 4.0 tech report. Source has no SimpleQA table.
exaone-4.0-32b,aime_2024,72.1,https://arxiv.org/abs/2504.21318,{},,,dropped,
exaone-4.0-32b,codeforces_rating,1650.0,https://llm-stats.com/benchmarks,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
exaone-4.0-32b,swe_bench_verified,45.0,https://llm-stats.com/benchmarks/swe-bench-verified,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
minimax-m2,mmlu_pro,82.0,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: minimax-m2=82.0.
minimax-m2,gpqa_diamond,77.7,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: minimax-m2=77.7.
minimax-m2,swe_bench_verified,69.4,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: minimax-m2=69.4.
minimax-m2,aa_intelligence_index,61.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2 model card: minimax-m2=61.
minimax-m2,hle,12.5,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: minimax-m2=12.5.
minimax-m2,aime_2025,78.3,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: minimax-m2=78.3.
minimax-m2,livecodebench,83.0,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: minimax-m2=83.0.
minimax-m2,browsecomp,44.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: minimax-m2=44.
minimax-m2,chatbot_arena_elo,1346.0,https://lmarena.ai/leaderboard/text,{},True,leaderboard,verified,lmarena.ai text arena (fetched 2026-04-29). Arena model name: minimax-m2 (rank 165). ELO updates continuously; score reflects latest available.
minimax-m2,humaneval,90.0,https://artificialanalysis.ai/articles/minimax-m2-benchmarks-and-analysis,{},,,needs_review,Reference URL is AA article page for minimax-m2. AA model page shows humaneval=null for minimax-m2. Cannot verify 90.0% from AA source. AA does not appear to have independently tested humaneval for this model.
minimax-m2,ifeval,88.0,https://artificialanalysis.ai/models/minimax-m2,{},,,needs_review,"Reference URL is AA model page (minimax-m2). AA uses field name ""ifbench"" (not ifeval); current value ifbench=0.723→72.3%. BP has 88.0% for ifeval. Major mismatch (15.7pp). Possible different benchmark/field or stale data. Current AA ifbench value: 72.3%."
minimax-m2,math_500,97.0,https://artificialanalysis.ai/models/minimax-m2,{},,,needs_review,"Reference URL is AA model page (minimax-m2), but AA shows math_500=null for minimax-m2. Cannot verify 97.0% from AA source."
minimax-m2,simpleqa,40.0,https://artificialanalysis.ai/models/minimax-m2,{},,,needs_review,Reference URL is AA model page (minimax-m2). AA does not publish SimpleQA scores; field is not present in AA model data for any model. Cannot verify from AA source. Original source needs identification.
minimax-m2,terminal_bench,30.0,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: minimax-m2=30.0.
minimax-m2,aime_2024,65.0,https://llm-stats.com/models/minimax-m2,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
minimax-m2,codeforces_rating,1700.0,https://llm-stats.com/models/minimax-m2,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
minimax-m2,mmlu,87.0,https://llm-stats.com/models/minimax-m2,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
minimax-m2,mmmu,75.0,https://llm-stats.com/models/minimax-m2,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
minimax-m2,swe_bench_pro,32.0,https://llm-stats.com/models/minimax-m2,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
minimax-m2,terminal_bench_1,42.0,https://www.tbench.ai/leaderboard/terminal-bench/1.0,{},,leaderboard,verified,"Confirmed 42.0% on tbench.ai Terminal-Bench 1.0 leaderboard (iFlow CLI scaffold, 2025-11-11)."
olmo-2-13b,mmlu,68.5,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct: mmlu=68.5. [R5d: prior unverified value 67.5 from https://allenai.org/blog/olmo2 deleted.]
olmo-2-13b,mmlu_pro,35.1,https://huggingface.co/allenai/OLMo-2-1124-13B,{},,,dropped,
olmo-2-13b,gpqa_diamond,25.0,https://allenai.org/blog/olmo2,{},,,dropped,
olmo-2-13b,humaneval,55.0,https://allenai.org/blog/olmo2,{},,,dropped,
olmo-2-13b,ifeval,82.6,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct: ifeval=82.6. [R5d: prior unverified value 68.0 from https://allenai.org/blog/olmo2 deleted.]
olmo-2-13b,aime_2024,5.0,https://allenai.org/blog/olmo3,{},,,dropped,
olmo-2-13b,livecodebench,18.0,https://allenai.org/blog/olmo3,{},,,dropped,
olmo-2-13b,math_500,38.0,https://allenai.org/blog/olmo3,{},,,dropped,
lfm2.5-1.2b-thinking,gpqa_diamond,37.86,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""default"",""harness"":""official (LFM eval)"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":0.05,""tools"":""none""}",True,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: lfm2.5-1.2b-thinking=37.86.
lfm2.5-1.2b-thinking,mmlu_pro,49.65,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""default"",""harness"":""official (LFM eval)"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":0.05,""tools"":""none""}",True,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: lfm2.5-1.2b-thinking=49.65.
lfm2.5-1.2b-thinking,aime_2025,31.73,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""default"",""harness"":""official (LFM eval)"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":0.05,""tools"":""none""}",True,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: lfm2.5-1.2b-thinking=31.73.
lfm2.5-1.2b-thinking,math_500,87.96,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""default"",""harness"":""official (LFM eval)"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":0.05,""tools"":""none""}",True,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: lfm2.5-1.2b-thinking=87.96.
lfm2.5-1.2b-thinking,ifeval,88.42,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""default"",""harness"":""official (LFM eval)"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":0.05,""tools"":""none""}",True,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: lfm2.5-1.2b-thinking=88.42.
phi-4-mini,mmlu,67.3,https://huggingface.co/microsoft/Phi-4-mini-instruct,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,"phi-4-mini HF model card: mmlu=67.3 (re-sourced from https://huggingface.co/microsoft/Phi-4-mini-instruct, matches BP value)"
phi-4-mini,gpqa_diamond,30.4,https://huggingface.co/microsoft/Phi-4-mini-instruct,{},,,dropped," | DROPPED: BP cell URL claims phi-4-mini HF or phi-4 HF, but gpqa_diamond value not in phi-4-mini HF model card. Ghost cell per R5g(a)."
phi-4-mini,humaneval,74.4,https://huggingface.co/microsoft/Phi-4-mini-instruct,{},,,dropped," | DROPPED: BP cell URL claims phi-4-mini HF or phi-4 HF, but humaneval value not in phi-4-mini HF model card. Ghost cell per R5g(a)."
phi-4-mini,mmlu_pro,52.8,https://huggingface.co/microsoft/Phi-4-mini-instruct,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,"phi-4-mini HF model card: mmlu_pro=52.8 (re-sourced from https://huggingface.co/microsoft/Phi-4-mini-instruct, matches BP value)"
phi-4-mini,gsm8k,88.6,https://huggingface.co/microsoft/Phi-4-mini-instruct,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,"phi-4-mini HF model card: gsm8k=88.6 (re-sourced from https://huggingface.co/microsoft/Phi-4-mini-instruct, matches BP value)"
phi-4-mini,ifeval,72.0,https://huggingface.co/microsoft/phi-4,{},,,dropped," | DROPPED: BP cell URL claims phi-4-mini HF or phi-4 HF, but ifeval value not in phi-4-mini HF model card. Ghost cell per R5g(a)."
phi-4-mini,livecodebench,28.0,https://huggingface.co/microsoft/phi-4,{},,,dropped," | DROPPED: BP cell URL claims phi-4-mini HF or phi-4 HF, but livecodebench value not in phi-4-mini HF model card. Ghost cell per R5g(a)."
phi-4-mini,math_500,75.0,https://huggingface.co/microsoft/phi-4,{},,,dropped," | DROPPED: BP cell URL claims phi-4-mini HF or phi-4 HF, but math_500 value not in phi-4-mini HF model card. Ghost cell per R5g(a)."
falcon3-10b,ifeval,78.17,https://huggingface.co/tiiuae/Falcon3-10B-Instruct,{},True,model_card,verified,Falcon3-10B-Instruct HF card shows IFEval=78.17.
falcon3-10b,mmlu,73.1,https://huggingface.co/blog/falcon3,{},,,dropped,
falcon3-10b,mmlu_pro,42.5,https://huggingface.co/blog/falcon3,{},,,dropped,
falcon3-10b,gpqa_diamond,30.0,https://huggingface.co/blog/falcon3,{},,,dropped,
falcon3-10b,humaneval,72.0,https://huggingface.co/blog/falcon3,{},,,dropped,
falcon3-10b,aime_2024,8.0,https://llm-stats.com/benchmarks,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
falcon3-10b,livecodebench,22.0,https://llm-stats.com/benchmarks,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
falcon3-10b,math_500,62.0,https://llm-stats.com/benchmarks,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
internlm3-8b,mmlu,76.6,https://github.com/InternLM/InternLM,"{""effort"":""default"",""harness"":""OpenCompass"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""tools"":""none""}",True,model_card,verified,GitHub README table: mmlu = 76.6. No * marker (non-thinking mode).
internlm3-8b,gpqa_diamond,37.4,https://github.com/InternLM/InternLM,"{""effort"":""default"",""harness"":""OpenCompass"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""tools"":""none""}",True,model_card,verified,GitHub README table: gpqa_diamond = 37.4. No * marker (non-thinking mode).
internlm3-8b,math_500,,https://github.com/InternLM/InternLM,"{""mode"":""thinking"",""notes"":""Thinking mode per footnote *""}",False,model_card,dropped,Dropped: GitHub README marks this 83.0 with * = Thinking Mode. BP internlm3-8b is non-thinking (is_reasoning=false). Wrong variant per R5g-b.
internlm3-8b,humaneval,82.3,https://github.com/InternLM/InternLM,"{""effort"":""default"",""harness"":""OpenCompass"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""tools"":""none""}",True,model_card,verified,GitHub README table: humaneval = 82.3. No * marker (non-thinking mode).
internlm3-8b,ifeval,79.3,https://github.com/InternLM/InternLM,"{""effort"":""default"",""harness"":""OpenCompass"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""tools"":""none""}",True,model_card,verified,GitHub README table: ifeval = 79.3. No * marker (non-thinking mode).
internlm3-8b,aime_2024,12.0,https://llm-stats.com/benchmarks,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
internlm3-8b,livecodebench,32.0,https://llm-stats.com/benchmarks,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
internlm3-8b,mmlu_pro,57.6,https://github.com/InternLM/InternLM,"{""effort"":""default"",""harness"":""OpenCompass"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""tools"":""none""}",True,model_card,verified,GitHub README: MMLU-Pro 0-shot = 57.6 (no *). Overwrote llm-stats value of 52.
seed-thinking-v1.5,gpqa_diamond,77.3,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: seed-thinking-v1.5=77.3.
seed-thinking-v1.5,aime_2024,86.7,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: seed-thinking-v1.5=86.7.
seed-thinking-v1.5,aime_2025,74.0,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: seed-thinking-v1.5=74.0.
seed-thinking-v1.5,livecodebench,64.9,https://github.com/ByteDance-Seed/Seed-Thinking-v1.5,{},,,dropped, [R5g-b] GitHub source reports LiveCodeBench v5 (64.9%); correct cell livecodebench_v5 already verified. Wrong benchmark_id column. Dropped.
seed-thinking-v1.5,humaneval,90.0,https://github.com/ByteDance-Seed/Seed-Thinking-v1.5,{},,,dropped, [R5g-a] GitHub README table for Seed-Thinking-v1.5 does not report HumanEval. Ghost cell — value not found in referenced source. Dropped.
seed-thinking-v1.5,ifeval,87.4,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: seed-thinking-v1.5=87.4.
seed-thinking-v1.5,math_500,95.0,https://github.com/ByteDance-Seed/Seed-Thinking-v1.5,{},,,dropped, [R5g-a] GitHub README table for Seed-Thinking-v1.5 does not report MATH-500. Ghost cell — value not found in referenced source. Dropped.
seed-thinking-v1.5,mmlu_pro,87.0,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: seed-thinking-v1.5=87.0.
seed-thinking-v1.5,hle,22.0,https://llm-stats.com/benchmarks,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
seed-thinking-v1.5,mmlu,87.0,https://llm-stats.com/benchmarks,{},,,dropped,"DROPPED (R5h): random third-party blog/aggregator, no primary trail"
seed-thinking-v1.5,simpleqa,12.9,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: seed-thinking-v1.5=12.9.
seed-thinking-v1.5,swe_bench_verified,47.0,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: seed-thinking-v1.5=47.0.
claude-mythos,mmmlu,92.7,https://www-cdn.anthropic.com/8b8380204f74670be75e81c820ca8dda846ab289.pdf,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Mythos sys card Table 6.3.A.
claude-opus-4.6,mmmlu,91.1,https://www-cdn.anthropic.com/8b8380204f74670be75e81c820ca8dda846ab289.pdf,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,"Mythos sys card Table 6.3.A: Opus 4.6 = 91.1%. Verified per Anthropic standard config (thinking max effort, default sampling, avg 5 trials)."
claude-opus-4.7,mmmlu,91.5,https://www.anthropic.com/news/claude-opus-4-7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""avg 14 non-English languages"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,Anthropic Opus 4.7 blog table: claude-opus-4.7 = 91.5%.
claude-mythos,graphwalks_bfs_256k_1m,80.0,https://www-cdn.anthropic.com/8b8380204f74670be75e81c820ca8dda846ab289.pdf,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Mythos sys card Table 6.3.A.
claude-opus-4.6,graphwalks_bfs_256k_1m,38.7,https://www-cdn.anthropic.com/8b8380204f74670be75e81c820ca8dda846ab289.pdf,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,"Mythos sys card Table 6.3.A: Opus 4.6 = 38.7%. Verified per Anthropic standard config (thinking max effort, default sampling, avg 5 trials)."
gpt-5.4,graphwalks_bfs_256k_1m,21.4,https://openai.com/index/introducing-gpt-5-4/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.4 blog table: 21.4.
claude-mythos,charxiv_reasoning,86.1,https://www-cdn.anthropic.com/8b8380204f74670be75e81c820ca8dda846ab289.pdf,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Mythos sys card Table 6.3.A: no-tools column. With-tools=93.2 not used.
claude-opus-4.6,charxiv_reasoning,61.5,https://www-cdn.anthropic.com/8b8380204f74670be75e81c820ca8dda846ab289.pdf,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,"Mythos sys card Table 6.3.A: Opus 4.6 = 61.5%. Verified per Anthropic standard config (thinking max effort, default sampling, avg 5 trials)."
claude-opus-4.7,charxiv_reasoning,82.1,https://www.anthropic.com/news/claude-opus-4-7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,Anthropic Opus 4.7 blog table: claude-opus-4.7 = 82.1%.
kimi-k2.6,charxiv_reasoning,80.4,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""max (98304 generation tokens)"",""harness"":""official (in-house SWE-agent-derived for SWE-Bench)"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@10 coding, avg@3 vision)"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,model_card,verified,Kimi K2.6 model card (HF moonshotai/Kimi-K2.6) Evaluation table: CharXiv RQ 80.4
claude-mythos,swe_bench_multilingual,87.3,https://www-cdn.anthropic.com/8b8380204f74670be75e81c820ca8dda846ab289.pdf,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic""}",True,tech_report,verified,Mythos sys card §6.4: 5 trials.
claude-opus-4.6,swe_bench_multilingual,77.8,https://www-cdn.anthropic.com/8b8380204f74670be75e81c820ca8dda846ab289.pdf,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic""}",True,tech_report,verified,"Mythos sys card Table 6.3.A: Opus 4.6 = 77.8%. Verified per Anthropic standard config (thinking max effort, default sampling, avg 5 trials)."
kimi-k2.6,swe_bench_multilingual,76.7,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): kimi-k2.6 (per their own canonical config) = 76.7.
deepseek-v4-pro,swe_bench_multilingual,76.2,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-pro (Max mode) = 76.2.
deepseek-v4-flash,swe_bench_multilingual,73.3,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-flash (Max mode) = 73.3.
claude-mythos,swe_bench_multimodal,59.0,https://www-cdn.anthropic.com/8b8380204f74670be75e81c820ca8dda846ab289.pdf,"{""effort"":""max"",""harness"":""Anthropic internal"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic""}",True,tech_report,verified,"Mythos sys card §6.4: internal harness; per benchmark canonical, lab-internal harness accepted."
claude-opus-4.6,swe_bench_multimodal,27.1,https://www-cdn.anthropic.com/8b8380204f74670be75e81c820ca8dda846ab289.pdf,"{""effort"":""max"",""harness"":""Anthropic internal"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic""}",True,tech_report,verified,"Mythos sys card Table 6.3.A: Opus 4.6 = 27.1%. Verified per Anthropic standard config (thinking max effort, default sampling, avg 5 trials)."
deepseek-v4-pro,hmmt_feb_2026,95.2,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-pro (Max mode) = 95.2.
deepseek-v4-flash,hmmt_feb_2026,94.8,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-flash (Max mode) = 94.8.
kimi-k2.6,hmmt_feb_2026,92.7,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""max (98304 generation tokens)"",""harness"":""official (in-house SWE-agent-derived for SWE-Bench)"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@10 coding, avg@3 vision)"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,model_card,verified,Kimi K2.6 model card (HF moonshotai/Kimi-K2.6) Evaluation table: HMMT 2026 Feb 92.7
gpt-5.4,hmmt_feb_2026,97.7,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): gpt-5.4 (per their own canonical config) = 97.7.
claude-opus-4.6,hmmt_feb_2026,96.2,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): claude-opus-4.6 (per their own canonical config) = 96.2.
gemini-3.1-pro,hmmt_feb_2026,94.7,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): gemini-3.1-pro (per their own canonical config) = 94.7.
kimi-k2.5,hmmt_feb_2026,81.3,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5.1 blog: kimi-k2.5=81.3.
glm-5.1,hmmt_feb_2026,82.6,https://z.ai/blog/glm-5.2,"{""context"":""default"",""effort"":""official reported"",""harness"":""official reasoning evaluation"",""judge"":""GPT-5.5 medium for AIME/HMMT/IMO; benchmark-specified otherwise"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none"",""top_p"":""0.95""}",True,official_blog,verified,GLM-5.2 official release source: GLM-5.1 / HMMT Feb. 2026 = 82.6.
deepseek-v4-pro,imo_answerbench,89.8,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-pro (Max mode) = 89.8.
deepseek-v4-flash,imo_answerbench,88.4,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-flash (Max mode) = 88.4.
kimi-k2.6,imo_answerbench,86.0,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): kimi-k2.6 (per their own canonical config) = 86.0.
gpt-5.4,imo_answerbench,91.4,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): gpt-5.4 (per their own canonical config) = 91.4.
claude-opus-4.6,imo_answerbench,75.3,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): claude-opus-4.6 (per their own canonical config) = 75.3.
kimi-k2.5,imo_answerbench,81.8,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default (96k for reasoning, 64k for vision)"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 AIME/HMMT, avg@8 GPQA, avg@3 vision, avg@4 Seal/WideSearch)"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,model_card,verified,Kimi K2.5 model card (HF moonshotai/Kimi-K2.5) Evaluation table: IMO-AnswerBench 81.8
glm-5.1,imo_answerbench,83.8,https://z.ai/blog/glm-5.2,"{""context"":""default"",""effort"":""official reported"",""harness"":""official reasoning evaluation"",""judge"":""GPT-5.5 medium for AIME/HMMT/IMO; benchmark-specified otherwise"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none"",""top_p"":""0.95""}",True,official_blog,verified,GLM-5.2 official release source: GLM-5.1 / IMOAnswerBench = 83.8.
deepseek-v4-pro,apex_shortlist,90.2,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-pro (Max mode) = 90.2.
deepseek-v4-flash,apex_shortlist,85.7,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-flash (Max mode) = 85.7.
kimi-k2.6,apex_shortlist,75.5,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): kimi-k2.6 (per their own canonical config) = 75.5.
gpt-5.4,apex_shortlist,78.1,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): gpt-5.4 (per their own canonical config) = 78.1.
claude-opus-4.6,apex_shortlist,85.9,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): claude-opus-4.6 (per their own canonical config) = 85.9.
gemini-3.1-pro,apex_shortlist,89.1,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): gemini-3.1-pro (per their own canonical config) = 89.1.
glm-5.1,apex_shortlist,72.4,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): glm-5.1 (per their own canonical config) = 72.4.
deepseek-v4-pro,mcpatlas,73.6,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-pro (Max mode) = 73.6.
deepseek-v4-flash,mcpatlas,69.0,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-flash (Max mode) = 69.0.
kimi-k2.6,mcpatlas,66.6,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",False,model_card,verified,"DeepSeek V4-Pro model card table: Kimi K2.6 Thinking scores 66.6 on MCPAtlas Public (Pass@1). This establishes the exact public benchmark variant and metric, but not the complete Kimi K2.6 max-effort, temperature, context, and harness setting."
gpt-5.4,mcpatlas,70.6,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 70.6.
gpt-5.5,mcpatlas,75.3,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 75.3.
claude-opus-4.6,mcpatlas,75.8,https://www.anthropic.com/news/claude-opus-4-7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,Anthropic Opus 4.7 blog table: claude-opus-4.6 = 75.8%.
claude-opus-4.7,mcpatlas,77.3,https://www.anthropic.com/news/claude-opus-4-7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,Anthropic Opus 4.7 blog table: claude-opus-4.7 = 77.3%.
gemini-3.1-pro,mcpatlas,69.2,https://deepmind.google/models/gemini/pro/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/pro/ Performance table (PROMOTED over prior third-party): MCP Atlas 69.2% (was 78.2 from OpenAI gpt-5.5 blog third-party; demoted to candidate)
glm-5.1,mcpatlas,71.8,https://z.ai/blog/glm-5.2,"{""context"":""default"",""effort"":""official reported"",""harness"":""public 500-task set; 10-minute timeout per task"",""judge"":""Gemini 3.0 Pro"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic MCP servers""}",False,official_blog,verified,GLM-5.2 official release source: GLM-5.1 / MCP-Atlas (Public Set) = 71.8.
deepseek-v4-pro,toolathlon,51.8,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-pro (Max mode) = 51.8.
deepseek-v4-flash,toolathlon,47.8,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-flash (Max mode) = 47.8.
kimi-k2.6,toolathlon,50.0,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): kimi-k2.6 (per their own canonical config) = 50.0.
gpt-5.4,toolathlon,54.6,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 54.6.
gpt-5.5,toolathlon,55.6,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 55.6.
claude-opus-4.6,toolathlon,47.2,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): claude-opus-4.6 (per their own canonical config) = 47.2.
gemini-3.1-pro,toolathlon,48.8,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): gemini-3.1-pro (per their own canonical config) = 48.8.
kimi-k2.5,toolathlon,27.8,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5.1 blog: kimi-k2.5=27.8.
glm-5.1,toolathlon,40.7,https://z.ai/blog/glm-5.2,"{""context"":""default"",""effort"":""official reported"",""harness"":""original Toolathlon public evaluation service (pre-Verified)"",""judge"":""official evaluation service"",""max_output_tokens"":""128000"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic tool use""}",True,official_blog,verified,GLM-5.2 official release source: GLM-5.1 / Tool-Decathlon = 40.7.
deepseek-v4-pro,simpleqa_verified,57.9,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-pro (Max mode) = 57.9.
deepseek-v4-flash,simpleqa_verified,34.1,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-flash (Max mode) = 34.1.
qwen3.5-397b,hmmt_feb_2026,87.88,https://huggingface.co/datasets/MathArena/hmmt_feb_2026,"""thinking:default""",True,leaderboard,verified,
deepseek-v3.2,hmmt_feb_2026,79.9,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5.1 blog: deepseek-v3.2=79.9.
qwen3-30b-a3b,hmmt_feb_2026,78.79,https://huggingface.co/datasets/MathArena/hmmt_feb_2026,{},,,dropped, | DROPPED: BP value from matharena 2026 dataset; BP model release_date=2025-05-15. Wrong-variant.
deepseek-v3.2,aime_2026,95.1,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5.1 blog: deepseek-v3.2=95.1.
qwen3-30b-a3b,aime_2026,87.5,https://huggingface.co/datasets/MathArena/aime_2026,{},,,dropped, | DROPPED: BP value from matharena 2026 dataset; BP model release_date=2025-05-15. Wrong-variant.
kimi-k2.5,aime_2026,94.5,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5.1 blog: kimi-k2.5=94.5.
gemma-4-31b,mmlu_pro,85.2,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-31b,aime_2026,89.2,https://blog.google/innovation-and-ai/technology/developers-tools/gemma-4/,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,official_blog,verified,"Google Gemma 4 family launch table, duplicated by the DeepMind semantic table."
gemma-4-31b,gpqa_diamond,84.3,https://blog.google/innovation-and-ai/technology/developers-tools/gemma-4/,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,official_blog,verified,"Google Gemma 4 family launch table, duplicated by the DeepMind semantic table."
gemma-4-31b,mmmlu,88.4,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-31b,codeforces_rating,2150.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""code execution in undisclosed harness""}",False,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-31b,hle,19.5,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-31b,mmmu_pro,76.9,https://blog.google/innovation-and-ai/technology/developers-tools/gemma-4/,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none"",""visual_tokens"":1120}",True,official_blog,verified,"Google Gemma 4 family launch table, duplicated by the DeepMind semantic table."
gemma-4-31b,mathvision,85.6,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none"",""visual_tokens"":1120}",True,tech_report,verified,"Gemma 4 Technical Report Table 6; maximum supported resolution, 1,120 visual tokens."
gemma-4-26b-a4b,mmlu_pro,82.6,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-26b-a4b,aime_2026,88.3,https://blog.google/innovation-and-ai/technology/developers-tools/gemma-4/,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,official_blog,verified,"Google Gemma 4 family launch table, duplicated by the DeepMind semantic table."
gemma-4-26b-a4b,gpqa_diamond,82.3,https://blog.google/innovation-and-ai/technology/developers-tools/gemma-4/,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,official_blog,verified,"Google Gemma 4 family launch table, duplicated by the DeepMind semantic table."
gemma-4-26b-a4b,mmmlu,86.3,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-26b-a4b,codeforces_rating,1718.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""code execution in undisclosed harness""}",False,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-26b-a4b,hle,8.7,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-26b-a4b,mmmu_pro,73.8,https://blog.google/innovation-and-ai/technology/developers-tools/gemma-4/,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none"",""visual_tokens"":1120}",True,official_blog,verified,"Google Gemma 4 family launch table, duplicated by the DeepMind semantic table."
gemma-4-26b-a4b,mathvision,82.4,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none"",""visual_tokens"":1120}",True,tech_report,verified,"Gemma 4 Technical Report Table 6; maximum supported resolution, 1,120 visual tokens."
gemma-4-e4b,mmlu_pro,69.4,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-e4b,aime_2026,42.5,https://blog.google/innovation-and-ai/technology/developers-tools/gemma-4/,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,official_blog,verified,"Google Gemma 4 family launch table, duplicated by the DeepMind semantic table."
gemma-4-e4b,gpqa_diamond,58.6,https://blog.google/innovation-and-ai/technology/developers-tools/gemma-4/,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,official_blog,verified,"Google Gemma 4 family launch table, duplicated by the DeepMind semantic table."
gemma-4-e4b,mmmlu,76.6,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-e4b,codeforces_rating,940.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""code execution in undisclosed harness""}",False,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-e4b,mmmu_pro,52.6,https://blog.google/innovation-and-ai/technology/developers-tools/gemma-4/,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none"",""visual_tokens"":1120}",True,official_blog,verified,"Google Gemma 4 family launch table, duplicated by the DeepMind semantic table."
gemma-4-e4b,mathvision,59.5,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none"",""visual_tokens"":1120}",True,tech_report,verified,"Gemma 4 Technical Report Table 6; maximum supported resolution, 1,120 visual tokens."
gemma-4-e2b,mmlu_pro,60.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-e2b,aime_2026,37.5,https://blog.google/innovation-and-ai/technology/developers-tools/gemma-4/,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,official_blog,verified,"Google Gemma 4 family launch table, duplicated by the DeepMind semantic table."
gemma-4-e2b,gpqa_diamond,43.4,https://blog.google/innovation-and-ai/technology/developers-tools/gemma-4/,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,official_blog,verified,"Google Gemma 4 family launch table, duplicated by the DeepMind semantic table."
gemma-4-e2b,mmmlu,67.4,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-e2b,codeforces_rating,633.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""code execution in undisclosed harness""}",False,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-e2b,mmmu_pro,44.2,https://blog.google/innovation-and-ai/technology/developers-tools/gemma-4/,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none"",""visual_tokens"":1120}",True,official_blog,verified,"Google Gemma 4 family launch table, duplicated by the DeepMind semantic table."
gemma-4-e2b,mathvision,52.4,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none"",""visual_tokens"":1120}",True,tech_report,verified,"Gemma 4 Technical Report Table 6; maximum supported resolution, 1,120 visual tokens."
gpt-5.4,usamo_2026,95.24,https://matharena.ai/?comp=usamo--usamo_2026,"""reasoning_effort:xhigh""",True,leaderboard,verified,
glm-5,hmmt_feb_2026,82.8,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5.1 blog: glm-5=82.8.
glm-5,aime_2026,95.4,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5.1 blog: glm-5=95.4.
glm-5,arc_agi_1,44.7,https://arcprize.org/leaderboard,"{""effort"":""default"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""CoT"",""tools"":""none""}",True,leaderboard,verified,arcprize.org leaderboard audit: GLM-5 on leaderboard
glm-5,arc_agi_2,4.9,https://arcprize.org/leaderboard,"{""effort"":""default"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""CoT"",""tools"":""none""}",True,leaderboard,verified,arcprize.org leaderboard audit: GLM-5 on leaderboard
glm-5,usamo_2026,35.12,https://matharena.ai/?comp=usamo--usamo_2026,"""thinking:default""",True,leaderboard,verified,
minimax-m2.5,arc_agi_1,63.7,https://arcprize.org/leaderboard,"{""effort"":""default"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""CoT"",""tools"":""none""}",True,leaderboard,verified,arcprize.org leaderboard audit: Minimax M2.5 on leaderboard
minimax-m2.5,arc_agi_2,4.9,https://arcprize.org/leaderboard,"{""effort"":""default"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""CoT"",""tools"":""none""}",True,leaderboard,verified,arcprize.org leaderboard audit: Minimax M2.5 on leaderboard
gemini-3.1-pro,usamo_2026,74.4,https://matharena.ai/?comp=usamo--usamo_2026,"""thinking:high""",True,leaderboard,verified,
claude-opus-4.6,usamo_2026,47.02,https://matharena.ai/?comp=usamo--usamo_2026,{},,,needs_review,matharena.ai/usamo/ currently shows only GPT-5.4 and Gemini models; claude-opus-4.6 entry no longer visible. Stale reference. R5g(c).
qwen3.5-397b,usamo_2026,36.31,https://matharena.ai/?comp=usamo--usamo_2026,"""thinking:default""",True,leaderboard,verified,
grok-3-beta,usamo_2025,4.76,https://matharena.ai/?comp=usamo--usamo_2025,{},,leaderboard,verified,
gemini-2.0-flash,usamo_2025,4.17,https://matharena.ai/?comp=usamo--usamo_2025,{},,,dropped,
claude-sonnet-4.6,bullshit_pushback,94.55,https://github.com/petergpt/bullshit-benchmark/tree/main/data/latest,"""leaderboard.csv green_rate\u00d7100; reasoning=none""",True,leaderboard,verified,GitHub repo source; not yet verified. R5g(c).
claude-opus-4.6,bullshit_pushback,92.73,https://github.com/petergpt/bullshit-benchmark/tree/main/data/latest,"""leaderboard.csv green_rate\u00d7100; reasoning=none""",True,leaderboard,verified,GitHub repo source; not yet verified. R5g(c).
claude-opus-4.5,bullshit_pushback,90.91,https://github.com/petergpt/bullshit-benchmark/tree/main/data/latest,"""leaderboard.csv green_rate\u00d7100; reasoning=high""",True,leaderboard,verified,GitHub repo source; not yet verified. R5g(c).
claude-haiku-4.5,bullshit_pushback,87.27,https://github.com/petergpt/bullshit-benchmark/tree/main/data/latest,"""leaderboard.csv green_rate\u00d7100; reasoning=high""",True,leaderboard,verified,GitHub repo source; not yet verified. R5g(c).
claude-opus-4.7,bullshit_pushback,80.0,https://github.com/petergpt/bullshit-benchmark/tree/main/data/latest,"""leaderboard.csv green_rate\u00d7100; reasoning=none""",True,leaderboard,verified,
grok-4.20,bullshit_pushback,67.27,https://github.com/petergpt/bullshit-benchmark/tree/main/data/latest,"""leaderboard.csv green_rate\u00d7100; reasoning=xhigh (multi-agent-beta variant)""",True,leaderboard,verified,
qwen3.5-397b,bullshit_pushback,65.45,https://github.com/petergpt/bullshit-benchmark/tree/main/data/latest,"""leaderboard.csv green_rate\u00d7100; reasoning=none""",True,leaderboard,verified,
claude-sonnet-4.5,bullshit_pushback,65.45,https://github.com/petergpt/bullshit-benchmark/tree/main/data/latest,"""leaderboard.csv green_rate\u00d7100; reasoning=high""",True,leaderboard,verified,GitHub repo source; not yet verified. R5g(c).
claude-3.7-sonnet,bullshit_pushback,54.72,https://github.com/petergpt/bullshit-benchmark/tree/main/data/latest,"""leaderboard.csv green_rate\u00d7100; reasoning=default, thinking variant""",True,leaderboard,verified,
llama-4-maverick,bullshit_pushback,54.55,,{},False,third_party,verified,BullshitPushback third-party benchmark.
kimi-k2.5,bullshit_pushback,47.27,https://github.com/petergpt/bullshit-benchmark/tree/main/data/latest,"""leaderboard.csv green_rate\u00d7100; reasoning=none""",True,leaderboard,verified,
claude-opus-4.1,bullshit_pushback,40.0,https://github.com/petergpt/bullshit-benchmark/tree/main/data/latest,"""leaderboard.csv green_rate\u00d7100; reasoning=none""",True,leaderboard,verified,GitHub repo source; not yet verified. R5g(c).
llama-4-scout,bullshit_pushback,36.36,,{},False,third_party,verified,BullshitPushback is third-party benchmark.
claude-opus-4,bullshit_pushback,32.73,https://github.com/petergpt/bullshit-benchmark/tree/main/data/latest,"""leaderboard.csv green_rate\u00d7100; reasoning=default""",True,leaderboard,verified,
gpt-5.2,bullshit_pushback,27.27,https://github.com/petergpt/bullshit-benchmark/tree/main/data/latest,"""leaderboard.csv green_rate\u00d7100; reasoning=none""",True,leaderboard,verified,
gpt-5.4,bullshit_pushback,25.45,https://github.com/petergpt/bullshit-benchmark/tree/main/data/latest,"""leaderboard.csv green_rate\u00d7100; reasoning=none""",True,leaderboard,verified,
claude-sonnet-4,bullshit_pushback,25.45,https://github.com/petergpt/bullshit-benchmark/tree/main/data/latest,"""leaderboard.csv green_rate\u00d7100; reasoning=none""",True,leaderboard,verified,
gpt-5.1,bullshit_pushback,25.45,https://github.com/petergpt/bullshit-benchmark/tree/main/data/latest,"""leaderboard.csv green_rate\u00d7100; reasoning=default""",True,leaderboard,verified,
gemini-2.5-pro,bullshit_pushback,23.64,https://github.com/petergpt/bullshit-benchmark/tree/main/data/latest,"""leaderboard.csv green_rate\u00d7100; reasoning=default""",True,leaderboard,verified,
gpt-5,bullshit_pushback,21.82,https://github.com/petergpt/bullshit-benchmark/tree/main/data/latest,"""leaderboard.csv green_rate\u00d7100; reasoning=default""",True,leaderboard,verified,
kimi-k2,bullshit_pushback,20.0,https://github.com/petergpt/bullshit-benchmark/tree/main/data/latest,{},,,dropped, | DROPPED (R5h-third-party-blog: community-benchmark)
gpt-5.3-codex,bullshit_pushback,14.55,https://github.com/petergpt/bullshit-benchmark/tree/main/data/latest,"""leaderboard.csv green_rate\u00d7100; reasoning=low""",True,leaderboard,verified,
gemini-2.5-flash,bullshit_pushback,14.55,https://github.com/petergpt/bullshit-benchmark/tree/main/data/latest,"""leaderboard.csv green_rate\u00d7100; reasoning=default""",True,leaderboard,verified,
deepseek-v3.2,bullshit_pushback,14.55,https://github.com/petergpt/bullshit-benchmark/tree/main/data/latest,"""leaderboard.csv green_rate\u00d7100; reasoning=high""",True,leaderboard,verified,
gpt-4.1,bullshit_pushback,14.55,https://github.com/petergpt/bullshit-benchmark/tree/main/data/latest,"""leaderboard.csv green_rate\u00d7100; reasoning=default""",True,leaderboard,verified,
deepseek-r1,bullshit_pushback,12.73,https://github.com/petergpt/bullshit-benchmark/tree/main/data/latest,"""leaderboard.csv green_rate\u00d7100; reasoning=none""",True,leaderboard,verified,
gpt-oss-120b,bullshit_pushback,3.64,https://github.com/petergpt/bullshit-benchmark/tree/main/data/latest,"""leaderboard.csv green_rate\u00d7100; reasoning=low""",True,leaderboard,verified,
kimi-k2.6,simpleqa_verified,36.9,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): kimi-k2.6 (per their own canonical config) = 36.9.
gpt-5.4,simpleqa_verified,45.3,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): gpt-5.4 (per their own canonical config) = 45.3.
claude-opus-4.6,simpleqa_verified,46.2,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): claude-opus-4.6 (per their own canonical config) = 46.2.
gemini-3.1-pro,simpleqa_verified,75.6,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): gemini-3.1-pro (per their own canonical config) = 75.6.
glm-5.1,simpleqa_verified,38.1,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): glm-5.1 (per their own canonical config) = 38.1.
amazon-nova-premier,aime_2025,16.0,https://cdn.amazon.science/e5/e6/ccc5378c42dca467d1abe1628ec9/amazon-nova-premier-technical-report-and-model-card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""zeroshot-cot"",""sampling"":""pass@1"",""temperature"":0.0,""tools"":""none""}",True,tech_report,verified,
amazon-nova-premier,charxiv_reasoning,48.8,https://cdn.amazon.science/e5/e6/ccc5378c42dca467d1abe1628ec9/amazon-nova-premier-technical-report-and-model-card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""zeroshot-cot"",""sampling"":""pass@1"",""temperature"":0.0,""tools"":""none""}",True,tech_report,verified,
amazon-nova-premier,simpleqa,86.3,https://cdn.amazon.science/e5/e6/ccc5378c42dca467d1abe1628ec9/amazon-nova-premier-technical-report-and-model-card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""zeroshot-cot"",""sampling"":""pass@1"",""temperature"":0.0,""tools"":""web""}",False,tech_report,verified,Reported with SerpApi web-search tool per Premier tech report; deviates from canonical tools=none.
amazon-nova-pro,aime_2025,5.3,https://cdn.amazon.science/e5/e6/ccc5378c42dca467d1abe1628ec9/amazon-nova-premier-technical-report-and-model-card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""zeroshot-cot"",""sampling"":""pass@1"",""temperature"":0.0,""tools"":""none""}",True,tech_report,verified,
amazon-nova-pro,mmmu,62.0,https://cdn.amazon.science/e5/e6/ccc5378c42dca467d1abe1628ec9/amazon-nova-premier-technical-report-and-model-card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""zeroshot-cot"",""sampling"":""pass@1"",""temperature"":0.0,""tools"":""none""}",True,tech_report,verified,
amazon-nova-pro,charxiv_reasoning,40.6,https://cdn.amazon.science/e5/e6/ccc5378c42dca467d1abe1628ec9/amazon-nova-premier-technical-report-and-model-card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""zeroshot-cot"",""sampling"":""pass@1"",""temperature"":0.0,""tools"":""none""}",True,tech_report,verified,
amazon-nova-pro,simpleqa,84.6,https://cdn.amazon.science/e5/e6/ccc5378c42dca467d1abe1628ec9/amazon-nova-premier-technical-report-and-model-card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""zeroshot-cot"",""sampling"":""pass@1"",""temperature"":0.0,""tools"":""web""}",False,tech_report,verified,Reported with SerpApi web-search tool per Premier tech report (cross-references Nova Pro); deviates from canonical tools=none.
claude-opus-4.6,usamo_2025,42.3,https://www-cdn.anthropic.com/8b8380204f74670be75e81c820ca8dda846ab289.pdf,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 10 trials)"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,"Mythos sys card Table 6.3.A: Opus 4.6 = 42.3%. Verified per Anthropic standard config (thinking max effort, default sampling, avg 5 trials)."
claude-opus-4,tau_bench_airline,59.6,https://www.anthropic.com/news/claude-4,"{""effort"":""default"",""harness"":""\u03c4-bench official + max steps 100"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""prompt addendum to Agent Policy (Anthropic)"",""sampling"":""pass@1 (single-pass)"",""temperature"":""top_p=0.95"",""tools"":""agentic (TAU-bench official tools)""}",False,official_blog,verified,Anthropic Claude 4 blog table. With extended thinking + tool use; prompt addendum to Airline Agent Policy; max steps 100.
claude-opus-4,mmmlu,88.8,https://www.anthropic.com/news/claude-4,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""avg 14 non-English languages"",""sampling"":""pass@1 (single-pass)"",""temperature"":""top_p=0.95"",""tools"":""none""}",True,official_blog,verified,"Anthropic Claude 4 blog table. Avg over 14 non-English languages, no extended thinking."
claude-sonnet-4,tau_bench_airline,60.0,https://www.anthropic.com/news/claude-4,"{""effort"":""default"",""harness"":""\u03c4-bench official + max steps 100"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""prompt addendum to Agent Policy (Anthropic)"",""sampling"":""pass@1 (single-pass)"",""temperature"":""top_p=0.95"",""tools"":""agentic (TAU-bench official tools)""}",False,official_blog,verified,Anthropic Claude 4 blog table. Prompt addendum + max steps 100.
claude-sonnet-4,mmmlu,86.5,https://www.anthropic.com/news/claude-4,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""avg 14 non-English languages"",""sampling"":""pass@1 (single-pass)"",""temperature"":""top_p=0.95"",""tools"":""none""}",True,official_blog,verified,Anthropic Claude 4 blog table. Avg over 14 non-English languages.
claude-3.7-sonnet,terminal_bench,35.2,https://www.anthropic.com/news/claude-4,"{""effort"":""default"",""harness"":""Claude Code agent framework"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (single-pass)"",""temperature"":""top_p=0.95"",""tools"":""agentic""}",True,official_blog,verified,Anthropic Claude 4 blog cross-model table: Sonnet 3.7 = 35.2%.
claude-opus-4.1,tau_bench_retail,82.4,https://www.anthropic.com/news/claude-opus-4-1,"{""effort"":""default"",""harness"":""\u03c4-bench official + max steps 100"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""prompt addendum to Agent Policy (Anthropic)"",""sampling"":""pass@1 (single-pass)"",""temperature"":""top_p=0.95"",""tools"":""agentic (TAU-bench official tools)""}",False,official_blog,verified,Anthropic Opus 4.1 blog table. Prompt addendum to Retail Agent Policy + max steps 100.
claude-opus-4.1,tau_bench_airline,56.0,https://www.anthropic.com/news/claude-opus-4-1,"{""effort"":""default"",""harness"":""\u03c4-bench official + max steps 100"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""prompt addendum to Agent Policy (Anthropic)"",""sampling"":""pass@1 (single-pass)"",""temperature"":""top_p=0.95"",""tools"":""agentic (TAU-bench official tools)""}",False,official_blog,verified,Anthropic Opus 4.1 blog table. Prompt addendum to Airline Agent Policy + max steps 100.
claude-opus-4.1,mmmu,77.1,https://www.anthropic.com/news/claude-opus-4-1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (single-pass)"",""temperature"":""top_p=0.95"",""tools"":""none""}",True,official_blog,verified,"Anthropic Opus 4.1 blog table (MMMU validation), with extended thinking."
claude-opus-4.1,mmmlu,89.5,https://www.anthropic.com/news/claude-opus-4-1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""avg 14 non-English languages"",""sampling"":""pass@1 (single-pass)"",""temperature"":""top_p=0.95"",""tools"":""none""}",True,official_blog,verified,"Anthropic Opus 4.1 blog table. Avg over 14 non-English languages, no extended thinking."
claude-opus-4.5,mmmlu,90.8,https://www.anthropic.com/news/claude-opus-4-5,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""avg 14 non-English languages"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,Anthropic Opus 4.5 blog table: 90.8% (avg 14 non-English languages).
claude-opus-4.5,tau2_bench_retail,88.9,https://www.anthropic.com/news/claude-opus-4-5,"{""effort"":""high"",""harness"":""\u03c4\u00b2-bench official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""per Anthropic, prompt addendum may apply"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic (\u03c4\u00b2-bench official tools)""}",False,official_blog,verified,Anthropic Opus 4.5 blog table: τ²-bench retail = 88.9%.
claude-opus-4.5,tau2_bench_telecom,98.2,https://www.anthropic.com/news/claude-opus-4-5,"{""effort"":""high"",""harness"":""\u03c4\u00b2-bench official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""per Anthropic, prompt addendum may apply"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic (\u03c4\u00b2-bench official tools)""}",False,official_blog,verified,Anthropic Opus 4.5 blog table: τ²-bench telecom = 98.2%.
claude-sonnet-4.5,mmmlu,89.1,https://www.anthropic.com/news/claude-opus-4-5,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""avg 14 non-English languages"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,Anthropic Opus 4.5 blog table: Sonnet 4.5 = 89.1%.
claude-opus-4.7,osworld,78.0,https://www.anthropic.com/news/claude-opus-4-7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,Anthropic Opus 4.7 blog table: claude-opus-4.7 = 78.0%.
claude-opus-4.7,finance_agent,64.4,https://www.anthropic.com/news/claude-opus-4-7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,Anthropic Opus 4.7 blog table: claude-opus-4.7 = 64.4%.
claude-opus-4.7,cybergym,73.1,https://www.anthropic.com/news/claude-opus-4-7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,Anthropic Opus 4.7 blog table: claude-opus-4.7 = 73.1%.
claude-opus-4.6,finance_agent,60.1,https://www.anthropic.com/news/claude-opus-4-7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,Anthropic Opus 4.7 blog table: claude-opus-4.6 = 60.1%.
claude-opus-4.6,cybergym,73.8,https://www.anthropic.com/news/claude-opus-4-7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,Anthropic Opus 4.7 blog table: claude-opus-4.6 = 73.8%.
claude-mythos,browsecomp,86.9,https://www.anthropic.com/news/claude-opus-4-7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,Anthropic Opus 4.7 blog table: claude-mythos = 86.9%.
claude-mythos,cybergym,83.1,https://www.anthropic.com/news/claude-opus-4-7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,Anthropic Opus 4.7 blog table: claude-mythos = 83.1%.
claude-sonnet-4.6,tau2_bench_retail,91.7,https://www.anthropic.com/news/claude-sonnet-4-6,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,Anthropic Sonnet 4.6 blog table: claude-sonnet-4.6 = 91.7.
claude-sonnet-4.6,tau2_bench_telecom,97.9,https://www.anthropic.com/news/claude-sonnet-4-6,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,Anthropic Sonnet 4.6 blog table: claude-sonnet-4.6 = 97.9.
claude-sonnet-4.6,mcpatlas,61.3,https://www.anthropic.com/news/claude-sonnet-4-6,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,Anthropic Sonnet 4.6 blog table: claude-sonnet-4.6 = 61.3.
claude-sonnet-4.6,finance_agent,63.3,https://www.anthropic.com/news/claude-sonnet-4-6,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,Anthropic Sonnet 4.6 blog table: claude-sonnet-4.6 = 63.3.
claude-sonnet-4.6,mmmlu,89.3,https://www.anthropic.com/news/claude-sonnet-4-6,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""avg 14 non-English languages"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,Anthropic Sonnet 4.6 blog table: claude-sonnet-4.6 = 89.3.
claude-sonnet-4.5,mcpatlas,43.8,https://www.anthropic.com/news/claude-sonnet-4-6,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,Anthropic Sonnet 4.6 blog table: claude-sonnet-4.5 = 43.8.
claude-sonnet-4.5,browsecomp,43.9,https://www.anthropic.com/news/claude-sonnet-4-6,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic (web search + fetch + programmatic tool calling, context compaction at 50k up to 10M)""}",True,official_blog,verified,Anthropic Sonnet 4.6 blog table: claude-sonnet-4.5 = 43.9.
claude-sonnet-4.5,finance_agent,54.5,https://www.anthropic.com/news/claude-sonnet-4-6,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,Anthropic Sonnet 4.6 blog table: claude-sonnet-4.5 = 54.5.
claude-sonnet-4.5,gdpval_aa_elo,1276.0,https://www.anthropic.com/news/claude-sonnet-4-6,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,Anthropic Sonnet 4.6 blog table: claude-sonnet-4.5 = 1276.
claude-opus-4.6,tau2_bench_retail,91.9,https://www.anthropic.com/news/claude-sonnet-4-6,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,Anthropic Sonnet 4.6 blog table: claude-opus-4.6 = 91.9.
claude-opus-4.6,tau2_bench_telecom,99.3,https://www.anthropic.com/news/claude-sonnet-4-6,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,Anthropic Sonnet 4.6 blog table: claude-opus-4.6 = 99.3.
claude-opus-4.5,mcpatlas,62.3,https://www.anthropic.com/news/claude-sonnet-4-6,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,Anthropic Sonnet 4.6 blog table: claude-opus-4.5 = 62.3.
claude-opus-4.5,finance_agent,58.8,https://www.anthropic.com/news/claude-sonnet-4-6,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,Anthropic Sonnet 4.6 blog table: claude-opus-4.5 = 58.8.
claude-opus-4.5,gdpval_aa_elo,1416.0,https://www.anthropic.com/news/claude-sonnet-4-6,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,Anthropic Sonnet 4.6 blog table: claude-opus-4.5 = 1416.
claude-opus-4.5,mmmu_pro,70.6,https://www.anthropic.com/news/claude-sonnet-4-6,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""updated grading: separate Sonnet 4 grader; no 'think step-by-step' prefix"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,Anthropic Sonnet 4.6 blog table: claude-opus-4.5 = 70.6.
gemini-3-pro,tau2_bench_retail,85.3,https://deepmind.google/models/gemini/pro/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/pro/ Performance table (PROMOTED over prior third-party): τ²-bench retail 85.3% (was 78.6 from Anthropic Sonnet 4.6 blog; demoted to candidate)
gemini-3-pro,tau2_bench_telecom,98.0,https://deepmind.google/models/gemini/pro/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/pro/ Performance table (PROMOTED over prior third-party): τ²-bench telecom 98.0% (was 89.2 from Anthropic Sonnet 4.6 blog; demoted to candidate)
gemini-3-pro,mmmlu,91.8,https://deepmind.google/models/gemini/pro/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/pro/ Performance table (PROMOTED over prior third-party): MMMLU 91.8% (was 92.0 from Anthropic Sonnet 4.6 blog; demoted to candidate)
gpt-5.2,mmmlu,89.6,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.2 blog: 89.6 (xhigh reasoning effort).
gpt-5.5,expert_swe,73.1,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 73.1.
gpt-5.5,gdpval_oai_woe,84.9,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 84.9.
gpt-5.5,osworld,78.7,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 78.7.
gpt-5.5,frontiermath_tier4,35.4,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 35.4.
gpt-5.5,cybergym,81.8,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 81.8.
gpt-5.5,ctf_internal,88.1,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 88.1.
gpt-5.5,finance_agent,60.0,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 60.0.
gpt-5.5,ib_modeling,88.5,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 88.5.
gpt-5.5,officeqa_pro,54.1,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 54.1.
gpt-5.5,bixbench,80.5,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 80.5.
gpt-5.5,genebench,25.0,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 25.0.
gpt-5.5,tau2_bench_telecom,98.0,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""original prompts (no prompt adjustment)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 98.0.
gpt-5.4,expert_swe,68.5,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 68.5.
gpt-5.4,gdpval_oai_woe,83.0,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 83.0.
gpt-5.4,osworld,75.0,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 75.0.
gpt-5.4,frontiermath_tier4,27.1,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 27.1.
gpt-5.4,cybergym,79.0,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 79.0.
gpt-5.4,ctf_internal,83.7,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 83.7.
gpt-5.4,finance_agent,56.0,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 56.0.
gpt-5.4,ib_modeling,87.3,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 87.3.
gpt-5.4,officeqa_pro,53.2,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 53.2.
gpt-5.4,bixbench,74.0,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 74.0.
gpt-5.4,genebench,19.0,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 19.0.
gpt-5.4,tau2_bench_telecom,92.8,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""original prompts (no prompt adjustment)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.5 blog table: 92.8.
gemini-3.1-pro,gdpval_oai_woe,67.3,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""unknown"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,third_party,verified,GPT-5.5 blog knowledge-work table: GDPval(wins or ties)=67.3%; third-party self-test by OpenAI
gemini-3.1-pro,frontiermath_tier4,16.7,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""unknown"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,third_party,verified,GPT-5.5 blog table: FrontierMath Tier 4=16.7%; third-party self-test by OpenAI
gemini-3.1-pro,finance_agent,59.7,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""unknown"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,third_party,verified,GPT-5.5 blog table: FinanceAgent v1.1=59.7%; third-party self-test by OpenAI
gemini-3.1-pro,officeqa_pro,18.1,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""unknown"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,third_party,verified,GPT-5.5 blog table: OfficeQA Pro=18.1%; third-party self-test by OpenAI
gpt-5.4,frontier_science_research,33.0,https://openai.com/index/introducing-gpt-5-4/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.4 blog table: 33.0.
gpt-5.4,officeqa,68.1,https://openai.com/index/introducing-gpt-5-4/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.4 blog table: 68.1.
gpt-5.4,omnidocbench,0.109,https://openai.com/index/introducing-gpt-5-4/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.4 blog table: 0.109.
gpt-5.4,graphwalks_bfs_0k_128k,93.0,https://openai.com/index/introducing-gpt-5-4/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.4 blog table: 93.0.
gpt-5.4,graphwalks_parents_0k_128k,89.8,https://openai.com/index/introducing-gpt-5-4/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.4 blog table: 89.8.
gpt-5.4,graphwalks_parents_256k_1m,32.4,https://openai.com/index/introducing-gpt-5-4/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.4 blog table: 32.4.
gpt-5.3-codex,gdpval_oai_woe,70.9,https://openai.com/index/introducing-gpt-5-3-codex/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.3-Codex blog: 70.9 (xhigh reasoning effort).
gpt-5.3-codex,toolathlon,51.9,https://openai.com/index/introducing-gpt-5-4/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.4 blog table: 51.9.
gpt-5.3-codex,browsecomp,77.3,https://openai.com/index/introducing-gpt-5-4/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.4 blog table: 77.3.
gpt-5.3-codex,finance_agent,54.0,https://openai.com/index/introducing-gpt-5-4/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.4 blog table: 54.0.
gpt-5.3-codex,ib_modeling,79.3,https://openai.com/index/introducing-gpt-5-4/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.4 blog table: 79.3.
gpt-5.3-codex,officeqa,65.1,https://openai.com/index/introducing-gpt-5-4/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.4 blog table: 65.1.
gpt-5.2,gdpval_oai_woe,70.9,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.2 blog: 70.9 (xhigh reasoning effort).
gpt-5.2,toolathlon,46.3,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.2 blog: 46.3 (xhigh reasoning effort).
gpt-5.2,mcpatlas,60.6,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.2 blog: 60.6 (xhigh reasoning effort).
gpt-5.2,finance_agent,59.5,https://openai.com/index/introducing-gpt-5-4/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.4 blog table: 59.5.
gpt-5.2,ib_modeling,68.4,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.2 blog: 68.4 (xhigh reasoning effort).
gpt-5.2,officeqa,63.1,https://openai.com/index/introducing-gpt-5-4/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.4 blog table: 63.1.
gpt-5.2,frontier_science_research,25.2,https://openai.com/index/introducing-gpt-5-4/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.4 blog table: 25.2.
gpt-5.2,frontiermath_tier4,14.6,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.2 blog: 14.6 (xhigh reasoning effort).
gpt-5.2,graphwalks_bfs_0k_128k,94.0,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.2 blog: 94.0 (xhigh reasoning effort).
gpt-5.2,graphwalks_parents_0k_128k,89.0,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.2 blog: 89.0 (xhigh reasoning effort).
gpt-5.2,omnidocbench,0.14,https://openai.com/index/introducing-gpt-5-4/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.4 blog table: 0.14.
gpt-5.2,tau2_bench_telecom,98.7,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""with prompt adjustment (Anthropic-style)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.2 blog: 98.7 (xhigh reasoning effort).
gpt-4.1,tau2_bench_telecom,34.0,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""with prompt adjustment (Anthropic-style)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1=34.0 (high reasoning effort).
gpt-5.3-codex,cybersecurity_ctf,77.6,https://openai.com/index/introducing-gpt-5-3-codex/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.3-Codex blog: 77.6 (xhigh reasoning effort).
gpt-5.3-codex,swelancer,81.4,https://openai.com/index/introducing-gpt-5-3-codex/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,OpenAI GPT-5.3-Codex blog: 81.4 (xhigh reasoning effort).
gpt-5.2,cybersecurity_ctf,67.7,https://openai.com/index/introducing-gpt-5-3-codex/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.3-Codex blog: 67.7 (xhigh reasoning effort).
gpt-5.2,swelancer,74.6,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,OpenAI GPT-5.2 blog: 74.6 (xhigh reasoning effort).
gpt-5.2,charxiv_reasoning,82.1,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (no-tools column)""}",True,official_blog,verified,OpenAI GPT-5.2 blog: 82.1 (xhigh reasoning effort).
gpt-5.2,tau2_bench_retail,82.0,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""with prompt adjustment (Anthropic-style)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.2 blog: 82.0 (xhigh reasoning effort).
gpt-5.2,hmmt_feb_2025,99.4,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.2 blog: 99.4 (xhigh reasoning effort).
gpt-5.1,swelancer,69.7,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,OpenAI GPT-5.2 blog (cross-model column): GPT-5.1 = 69.7.
gpt-5.1,charxiv_reasoning,67.0,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (no-tools column)""}",True,official_blog,verified,OpenAI GPT-5.2 blog (cross-model column): GPT-5.1 = 67.0.
gpt-5.1,video_mmmu,82.9,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.2 blog (cross-model column): GPT-5.1 = 82.9.
gpt-5.1,tau2_bench_telecom,95.6,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""with prompt adjustment (Anthropic-style)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.2 blog (cross-model column): GPT-5.1 = 95.6.
gpt-5.1,tau2_bench_retail,77.9,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""with prompt adjustment (Anthropic-style)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.2 blog (cross-model column): GPT-5.1 = 77.9.
gpt-5.1,mcpatlas,44.5,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.2 blog (cross-model column): GPT-5.1 = 44.5.
gpt-5.1,toolathlon,36.1,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.2 blog (cross-model column): GPT-5.1 = 36.1.
gpt-5.1,mmmlu,89.5,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.2 blog (cross-model column): GPT-5.1 = 89.5.
gpt-5.1,hmmt_feb_2025,96.3,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.2 blog (cross-model column): GPT-5.1 = 96.3.
gpt-5.1,frontiermath_tier4,12.5,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.2 blog (cross-model column): GPT-5.1 = 12.5.
gpt-5.1,graphwalks_bfs_0k_128k,76.8,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.2 blog (cross-model column): GPT-5.1 = 76.8.
gpt-5.1,graphwalks_parents_0k_128k,71.5,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5.2 blog (cross-model column): GPT-5.1 = 71.5.
gpt-5.1,ib_modeling,59.1,https://openai.com/index/introducing-gpt-5-2/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.2 blog (cross-model column): GPT-5.1 = 59.1.
gpt-5.1,tau2_bench_airline,67.0,https://openai.com/index/gpt-5-1-for-developers/,"{""effort"":""high"",""harness"":""official with apply_patch"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.1 dev blog: 67.0 (high reasoning effort).
gpt-5.1,browsecomp_long_context_128k,90.0,https://openai.com/index/gpt-5-1-for-developers/,"{""effort"":""high"",""harness"":""official with apply_patch"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.1 dev blog: 90.0 (high reasoning effort).
gpt-5,browsecomp_long_context_128k,90.0,https://openai.com/index/gpt-5-1-for-developers/,"{""effort"":""high"",""harness"":""official with apply_patch"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5.1 dev blog (cross-model column): GPT-5 = 90.0.
gpt-5,aider_polyglot_diff,88.0,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (diff)""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-5=88.0 (high reasoning effort).
gpt-5,collie,99.0,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-5=99.0 (high reasoning effort).
gpt-5,erqa,65.7,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-5=65.7 (high reasoning effort).
gpt-5,factscore,2.8,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-5=2.8 (high reasoning effort).
gpt-5,graphwalks_bfs_0k_128k,78.3,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-5=78.3 (high reasoning effort).
gpt-5,graphwalks_parents_0k_128k,73.3,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-5=73.3 (high reasoning effort).
gpt-5,internal_api_if_hard,64.0,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-5=64.0 (high reasoning effort).
gpt-5,longfact_concepts,1.0,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-5=1.0 (high reasoning effort).
gpt-5,longfact_objects,1.2,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-5=1.2 (high reasoning effort).
gpt-5,mmmu_pro,78.4,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""avg across standard and vision sets"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-5=78.4 (high reasoning effort).
gpt-5,multichallenge_o3mini_grader,69.6,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-5=69.6 (high reasoning effort).
gpt-5,swelancer_freelance_dollars,112000.0,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-5=112000 (high reasoning effort).
gpt-5,video_mmmu,84.6,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-5=84.6 (high reasoning effort).
gpt-5,video_mme,86.7,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-5=86.7 (high reasoning effort).
gpt-5,charxiv_reasoning,81.1,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""with python tool (per OpenAI dev blog)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""python tool""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-5=81.1 (high reasoning effort).
o3-high,aider_polyglot_diff,79.6,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (diff)""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o3-high=79.6 (high reasoning effort).
o3-high,browsecomp_long_context_128k,88.3,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o3-high=88.3 (high reasoning effort).
o3-high,collie,98.4,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o3-high=98.4 (high reasoning effort).
o3-high,erqa,64.0,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o3-high=64.0 (high reasoning effort).
o3-high,factscore,23.5,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o3-high=23.5 (high reasoning effort).
o3-high,graphwalks_bfs_0k_128k,77.3,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o3-high=77.3 (high reasoning effort).
o3-high,graphwalks_parents_0k_128k,72.9,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o3-high=72.9 (high reasoning effort).
o3-high,internal_api_if_hard,47.4,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o3-high=47.4 (high reasoning effort).
o3-high,longfact_concepts,5.2,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o3-high=5.2 (high reasoning effort).
o3-high,longfact_objects,6.8,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o3-high=6.8 (high reasoning effort).
o3-high,multichallenge_o3mini_grader,60.4,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o3-high=60.4 (high reasoning effort).
o3-high,swelancer_freelance_dollars,86000.0,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o3-high=86000 (high reasoning effort).
o3-high,tau2_bench_airline,64.8,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""with prompt adjustment (Anthropic-style)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o3-high=64.8 (high reasoning effort).
o3-high,tau2_bench_retail,80.2,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""with prompt adjustment (Anthropic-style)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o3-high=80.2 (high reasoning effort).
o3-high,tau2_bench_telecom,58.2,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""with prompt adjustment (Anthropic-style)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o3-high=58.2 (high reasoning effort).
o3-high,video_mmmu,83.3,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o3-high=83.3 (high reasoning effort).
o3-high,video_mme,84.9,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o3-high=84.9 (high reasoning effort).
o3-high,charxiv_reasoning,78.6,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""with python tool (per OpenAI dev blog)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""python tool""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o3-high=78.6 (high reasoning effort).
o4-mini-high,aider_polyglot_diff,58.2,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (diff)""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o4-mini-high=58.2 (high reasoning effort).
o4-mini-high,browsecomp_long_context_128k,80.0,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o4-mini-high=80.0 (high reasoning effort).
o4-mini-high,collie,96.1,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o4-mini-high=96.1 (high reasoning effort).
o4-mini-high,erqa,56.5,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o4-mini-high=56.5 (high reasoning effort).
o4-mini-high,factscore,38.7,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o4-mini-high=38.7 (high reasoning effort).
o4-mini-high,graphwalks_bfs_0k_128k,62.3,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o4-mini-high=62.3 (high reasoning effort).
o4-mini-high,graphwalks_parents_0k_128k,51.1,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o4-mini-high=51.1 (high reasoning effort).
o4-mini-high,internal_api_if_hard,44.7,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o4-mini-high=44.7 (high reasoning effort).
o4-mini-high,longfact_concepts,3.0,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o4-mini-high=3.0 (high reasoning effort).
o4-mini-high,longfact_objects,8.9,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o4-mini-high=8.9 (high reasoning effort).
o4-mini-high,mmmu_pro,73.4,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""avg across standard and vision sets"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o4-mini-high=73.4 (high reasoning effort).
o4-mini-high,multichallenge_o3mini_grader,57.5,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o4-mini-high=57.5 (high reasoning effort).
o4-mini-high,swelancer_freelance_dollars,66000.0,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o4-mini-high=66000 (high reasoning effort).
o4-mini-high,tau2_bench_airline,60.2,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""with prompt adjustment (Anthropic-style)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o4-mini-high=60.2 (high reasoning effort).
o4-mini-high,tau2_bench_retail,70.5,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""with prompt adjustment (Anthropic-style)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o4-mini-high=70.5 (high reasoning effort).
o4-mini-high,tau2_bench_telecom,40.5,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""with prompt adjustment (Anthropic-style)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o4-mini-high=40.5 (high reasoning effort).
o4-mini-high,video_mmmu,79.4,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o4-mini-high=79.4 (high reasoning effort).
o4-mini-high,video_mme,79.5,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o4-mini-high=79.5 (high reasoning effort).
o4-mini-high,charxiv_reasoning,72.0,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""with python tool (per OpenAI dev blog)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""python tool""}",True,official_blog,verified,OpenAI GPT-5 dev blog: o4-mini-high=72.0 (high reasoning effort).
gpt-4.1,aider_polyglot_diff,52.9,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (diff)""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1=52.9 (high reasoning effort).
gpt-4.1,browsecomp_long_context_128k,85.9,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1=85.9 (high reasoning effort).
gpt-4.1,collie,65.8,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1=65.8 (high reasoning effort).
gpt-4.1,erqa,44.3,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1=44.3 (high reasoning effort).
gpt-4.1,factscore,6.7,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1=6.7 (high reasoning effort).
gpt-4.1,graphwalks_bfs_0k_128k,61.7,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1=61.7 (high reasoning effort).
gpt-4.1,graphwalks_parents_0k_128k,58.0,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1=58.0 (high reasoning effort).
gpt-4.1,internal_api_if_hard,49.1,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1=49.1 (high reasoning effort).
gpt-4.1,longfact_concepts,0.7,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1=0.7 (high reasoning effort).
gpt-4.1,longfact_objects,1.1,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1=1.1 (high reasoning effort).
gpt-4.1,mmmu_pro,60.3,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""avg across standard and vision sets"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1=60.3 (high reasoning effort).
gpt-4.1,multichallenge_o3mini_grader,46.2,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1=46.2 (high reasoning effort).
gpt-4.1,swelancer_freelance_dollars,34000.0,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1=34000 (high reasoning effort).
gpt-4.1,tau2_bench_airline,56.0,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""with prompt adjustment (Anthropic-style)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1=56.0 (high reasoning effort).
gpt-4.1,tau2_bench_retail,74.0,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""with prompt adjustment (Anthropic-style)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1=74.0 (high reasoning effort).
gpt-4.1,video_mmmu,60.9,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1=60.9 (high reasoning effort).
gpt-4.1,video_mme,78.7,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1=78.7 (high reasoning effort).
gpt-4.1,charxiv_reasoning,56.7,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""with python tool (per OpenAI dev blog)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""python tool""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1=56.7 (high reasoning effort).
gpt-4.1-mini,aider_polyglot_diff,31.6,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (diff)""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-mini=31.6 (high reasoning effort).
gpt-4.1-mini,aime_2025,40.2,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-mini=40.2 (high reasoning effort).
gpt-4.1-mini,browsecomp_long_context_128k,89.0,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-mini=89.0 (high reasoning effort).
gpt-4.1-mini,collie,54.6,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-mini=54.6 (high reasoning effort).
gpt-4.1-mini,erqa,42.3,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-mini=42.3 (high reasoning effort).
gpt-4.1-mini,factscore,10.9,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-mini=10.9 (high reasoning effort).
gpt-4.1-mini,graphwalks_bfs_0k_128k,61.7,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-mini=61.7 (high reasoning effort).
gpt-4.1-mini,graphwalks_parents_0k_128k,60.5,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-mini=60.5 (high reasoning effort).
gpt-4.1-mini,hle,3.7,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-mini=3.7 (high reasoning effort).
gpt-4.1-mini,hmmt_feb_2025,35.0,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-mini=35.0 (high reasoning effort).
gpt-4.1-mini,internal_api_if_hard,45.1,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-mini=45.1 (high reasoning effort).
gpt-4.1-mini,longfact_concepts,1.1,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-mini=1.1 (high reasoning effort).
gpt-4.1-mini,longfact_objects,1.8,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-mini=1.8 (high reasoning effort).
gpt-4.1-mini,mmmu_pro,58.9,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""avg across standard and vision sets"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-mini=58.9 (high reasoning effort).
gpt-4.1-mini,multichallenge_o3mini_grader,42.2,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-mini=42.2 (high reasoning effort).
gpt-4.1-mini,swe_bench_verified,23.6,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,dropped,"OpenAI GPT-5 dev blog: gpt-4.1-mini=23.6 (high reasoning effort). DROPPED: source uses 477/500 subset (omit 23 problems), not full SWE-bench Verified."
gpt-4.1-mini,swelancer_freelance_dollars,31000.0,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-mini=31000 (high reasoning effort).
gpt-4.1-mini,tau2_bench_airline,51.0,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""with prompt adjustment (Anthropic-style)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-mini=51.0 (high reasoning effort).
gpt-4.1-mini,tau2_bench_retail,66.0,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""with prompt adjustment (Anthropic-style)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-mini=66.0 (high reasoning effort).
gpt-4.1-mini,tau2_bench_telecom,44.0,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""with prompt adjustment (Anthropic-style)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-mini=44.0 (high reasoning effort).
gpt-4.1-mini,video_mmmu,55.1,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-mini=55.1 (high reasoning effort).
gpt-4.1-mini,video_mme,68.4,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-mini=68.4 (high reasoning effort).
gpt-4.1-mini,charxiv_reasoning,56.8,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""with python tool (per OpenAI dev blog)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""python tool""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-mini=56.8 (high reasoning effort).
gpt-4.1-nano,aider_polyglot_diff,6.2,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (diff)""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-nano=6.2 (high reasoning effort).
gpt-4.1-nano,browsecomp_long_context_128k,89.4,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-nano=89.4 (high reasoning effort).
gpt-4.1-nano,collie,42.5,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-nano=42.5 (high reasoning effort).
gpt-4.1-nano,erqa,26.5,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-nano=26.5 (high reasoning effort).
gpt-4.1-nano,graphwalks_bfs_0k_128k,25.0,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-nano=25.0 (high reasoning effort).
gpt-4.1-nano,graphwalks_parents_0k_128k,9.4,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-nano=9.4 (high reasoning effort).
gpt-4.1-nano,internal_api_if_hard,31.6,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-nano=31.6 (high reasoning effort).
gpt-4.1-nano,mmmu,55.4,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-nano=55.4 (high reasoning effort).
gpt-4.1-nano,mmmu_pro,33.0,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""avg across standard and vision sets"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-nano=33.0 (high reasoning effort).
gpt-4.1-nano,multichallenge_o3mini_grader,31.1,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-nano=31.1 (high reasoning effort).
gpt-4.1-nano,swelancer_freelance_dollars,9000.0,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-nano=9000 (high reasoning effort).
gpt-4.1-nano,tau2_bench_airline,14.0,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""with prompt adjustment (Anthropic-style)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-nano=14.0 (high reasoning effort).
gpt-4.1-nano,tau2_bench_retail,21.5,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""with prompt adjustment (Anthropic-style)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-nano=21.5 (high reasoning effort).
gpt-4.1-nano,tau2_bench_telecom,12.1,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""with prompt adjustment (Anthropic-style)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-nano=12.1 (high reasoning effort).
gpt-4.1-nano,video_mmmu,30.2,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-nano=30.2 (high reasoning effort).
gpt-4.1-nano,video_mme,55.2,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-nano=55.2 (high reasoning effort).
gpt-4.1-nano,charxiv_reasoning,40.5,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""with python tool (per OpenAI dev blog)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""python tool""}",True,official_blog,verified,OpenAI GPT-5 dev blog: gpt-4.1-nano=40.5 (high reasoning effort).
gemini-3-pro,livecodebench_pro,2439.0,https://deepmind.google/models/gemini/pro/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/pro/ Performance table: LiveCodeBench Pro Elo 2439
gemini-3-pro,apex_agents,18.4,https://deepmind.google/models/gemini/pro/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/pro/ Performance table: APEX-Agents 18.4% (long-horizon professional tasks)
gemini-3-pro,gdpval_aa_elo,1195.0,https://deepmind.google/models/gemini/pro/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/pro/ Performance table: GDPval-AA Elo 1195
gemini-3-pro,mcpatlas,54.1,https://deepmind.google/models/gemini/pro/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/pro/ Performance table: MCP Atlas 54.1%
gemini-3.1-pro,tau2_bench_retail,90.8,https://deepmind.google/models/gemini/pro/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/pro/ Performance table: τ²-bench retail 90.8% (model card had earlier 90.5 — added as candidate)
gemini-3.1-pro,tau2_bench_telecom,99.3,https://deepmind.google/models/gemini/pro/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/pro/ Performance table: τ²-bench telecom 99.3%
gemini-3.1-pro,mmmlu,92.6,https://deepmind.google/models/gemini/pro/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/pro/ Performance table: MMMLU 92.6%
gemini-3.1-pro,apex_agents,33.5,https://deepmind.google/models/gemini/pro/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/pro/ Performance table: APEX-Agents 33.5%
gemini-3.1-pro,livecodebench_pro,2887.0,https://deepmind.google/models/gemini/pro/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/pro/ Performance table: LiveCodeBench Pro Elo 2887
claude-sonnet-4.6,scicode,47.0,https://deepmind.google/models/gemini/pro/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,Created from https://deepmind.google/models/gemini/pro/ (cross-model table). DeepMind /models/gemini/pro/ cross-model Performance table: SciCode 47% [Promoted to verified — official DeepMind table.]
claude-opus-4.6,scicode,52.0,https://deepmind.google/models/gemini/pro/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,Created from https://deepmind.google/models/gemini/pro/ (cross-model table). DeepMind /models/gemini/pro/ cross-model Performance table: SciCode 52% [Promoted to verified — official DeepMind table.]
claude-opus-4.6,apex_agents,29.8,https://deepmind.google/models/gemini/pro/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,Created from https://deepmind.google/models/gemini/pro/ (cross-model table). DeepMind /models/gemini/pro/ cross-model Performance table: APEX-Agents 29.8% [Promoted to verified — official DeepMind table.]
gpt-5.2,livecodebench_pro,2393.0,https://deepmind.google/models/gemini/pro/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,Created from https://deepmind.google/models/gemini/pro/ (cross-model table). DeepMind /models/gemini/pro/ cross-model Performance table: LiveCodeBench Pro Elo 2393 [Promoted to verified — official DeepMind table.]
gpt-5.2,apex_agents,23.0,https://deepmind.google/models/gemini/pro/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,Created from https://deepmind.google/models/gemini/pro/ (cross-model table). DeepMind /models/gemini/pro/ cross-model Performance table: APEX-Agents 23.0% [Promoted to verified — official DeepMind table.]
gpt-5.2,gdpval_aa_elo,1462.0,https://deepmind.google/models/gemini/pro/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,Created from https://deepmind.google/models/gemini/pro/ (cross-model table). DeepMind /models/gemini/pro/ cross-model Performance table: GDPval-AA Elo 1462 [Promoted to verified — official DeepMind table.]
gemini-3-flash,screenspot_pro,69.1,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: ScreenSpot-Pro no tools 69.1%
gemini-3-flash,charxiv_reasoning,80.3,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: CharXiv Reasoning no tools 80.3%
gemini-3-flash,omnidocbench,0.121,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: OmniDocBench 1.5 overall edit distance 0.121 (lower is better)
gemini-3-flash,livecodebench_pro,2316.0,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: LiveCodeBench Pro Elo 2316
gemini-3-flash,toolathlon,49.4,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: Toolathlon 49.4%
gemini-3-flash,mcpatlas,57.4,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: MCP Atlas 57.4%
gemini-3-flash,vending_bench_2,3635.0,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: Vending-Bench 2 net worth $3635 (mean)
gemini-3-flash,facts_benchmark,61.9,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: FACTS Benchmark Suite 61.9%
gemini-3-flash,simpleqa_verified,68.7,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: SimpleQA Verified 68.7%
gemini-3-flash,global_piqa,92.8,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: Global PIQA 92.8%
gemini-3-flash,mmmlu,91.8,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: MMMLU 91.8%
gemini-3-flash,mrcr_v2,67.2,https://deepmind.google/models/gemini/flash/,"{""context"":""128k"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: MRCR v2 8-needle 128k average 67.2%
gemini-3-pro,screenspot_pro,72.7,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: ScreenSpot-Pro 72.7%
gemini-3-pro,charxiv_reasoning,81.4,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: CharXiv Reasoning 81.4%
gemini-3-pro,omnidocbench,0.115,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: OmniDocBench 1.5 edit distance 0.115
gemini-3-pro,toolathlon,36.4,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: Toolathlon 36.4%
gemini-3-pro,vending_bench_2,5478.0,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: Vending-Bench 2 net worth $5478 (mean)
gemini-3-pro,facts_benchmark,70.5,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: FACTS Benchmark Suite 70.5%
gemini-3-pro,simpleqa_verified,72.1,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: SimpleQA Verified 72.1%
gemini-3-pro,global_piqa,93.4,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: Global PIQA 93.4%
gemini-2.5-flash,mmmu_pro,66.7,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: MMMU-Pro 66.7%
gemini-2.5-flash,screenspot_pro,3.9,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: ScreenSpot-Pro 3.9%
gemini-2.5-flash,charxiv_reasoning,63.7,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: CharXiv Reasoning 63.7%
gemini-2.5-flash,omnidocbench,0.154,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: OmniDocBench 1.5 edit distance 0.154
gemini-2.5-flash,livecodebench_pro,1143.0,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: LiveCodeBench Pro Elo 1143
gemini-2.5-flash,toolathlon,3.7,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: Toolathlon 3.7%
gemini-2.5-flash,mcpatlas,3.4,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: MCP Atlas 3.4%
gemini-2.5-flash,vending_bench_2,549.0,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: Vending-Bench 2 net worth $549
gemini-2.5-flash,facts_benchmark,50.4,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: FACTS Benchmark Suite 50.4%
gemini-2.5-flash,simpleqa_verified,28.1,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: SimpleQA Verified 28.1%
gemini-2.5-flash,mmmlu,86.6,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: MMMLU 86.6%
gemini-2.5-flash,global_piqa,90.2,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: Global PIQA 90.2%
gemini-2.5-flash,mrcr_v2,54.3,https://deepmind.google/models/gemini/flash/,"{""context"":""128k"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: MRCR v2 8-needle 128k 54.3%
gemini-2.5-pro,mmmu_pro,68.0,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: MMMU-Pro 68.0%
gemini-2.5-pro,screenspot_pro,11.4,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: ScreenSpot-Pro 11.4%
gemini-2.5-pro,charxiv_reasoning,69.6,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: CharXiv Reasoning 69.6%
gemini-2.5-pro,omnidocbench,0.145,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: OmniDocBench 1.5 edit distance 0.145
gemini-2.5-pro,livecodebench_pro,1775.0,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: LiveCodeBench Pro Elo 1775
gemini-2.5-pro,toolathlon,10.5,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: Toolathlon 10.5%
gemini-2.5-pro,mcpatlas,8.8,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: MCP Atlas 8.8%
gemini-2.5-pro,vending_bench_2,574.0,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: Vending-Bench 2 net worth $574
gemini-2.5-pro,facts_benchmark,63.4,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: FACTS Benchmark Suite 63.4%
gemini-2.5-pro,simpleqa_verified,54.5,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: SimpleQA Verified 54.5%
gemini-2.5-pro,mmmlu,89.5,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: MMMLU 89.5%
gemini-2.5-pro,global_piqa,91.5,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: Global PIQA 91.5%
claude-sonnet-4.5,screenspot_pro,36.2,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: ScreenSpot-Pro 36.2%
claude-sonnet-4.5,charxiv_reasoning,68.5,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: CharXiv Reasoning 68.5%
claude-sonnet-4.5,omnidocbench,0.145,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: OmniDocBench 1.5 edit distance 0.145
claude-sonnet-4.5,livecodebench_pro,1418.0,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: LiveCodeBench Pro Elo 1418
claude-sonnet-4.5,toolathlon,38.9,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: Toolathlon 38.9%
claude-sonnet-4.5,vending_bench_2,3839.0,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: Vending-Bench 2 net worth $3839
claude-sonnet-4.5,facts_benchmark,48.9,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: FACTS Benchmark Suite 48.9%
claude-sonnet-4.5,simpleqa_verified,29.3,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: SimpleQA Verified 29.3%
claude-sonnet-4.5,global_piqa,90.1,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: Global PIQA 90.1%
gpt-5.2,screenspot_pro,86.3,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""code""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: ScreenSpot-Pro with python 86.3%
gpt-5.2,vending_bench_2,3952.0,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: Vending-Bench 2 net worth $3952
gpt-5.2,facts_benchmark,61.4,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: FACTS Benchmark Suite 61.4%
gpt-5.2,simpleqa_verified,38.0,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: SimpleQA Verified 38.0%
gpt-5.2,global_piqa,91.2,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: Global PIQA 91.2%
grok-4.1,facts_benchmark,42.1,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: FACTS Benchmark Suite 42.1%
grok-4.1,simpleqa_verified,19.5,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: SimpleQA Verified 19.5%
grok-4.1,mmmlu,86.8,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: MMMLU 86.8%
grok-4.1,global_piqa,85.6,https://deepmind.google/models/gemini/flash/,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: Global PIQA 85.6%
grok-4.1,mrcr_v2,54.6,https://deepmind.google/models/gemini/flash/,"{""context"":""128k"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: MRCR v2 8-needle 128k 54.6%
gemini-2.5-pro,facts_grounding,87.8,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-5-Pro-Model-Card.pdf,"{""effort"":""default"",""harness"":""AI Studio API"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (single attempt)"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.5 Pro Model Card (June 2025 GA): FACTS Grounding 87.8%
gemini-2.5-pro,video_mme,86.9,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-5-Pro-Model-Card.pdf,"{""effort"":""default"",""harness"":""AI Studio API"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (single attempt)"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.5 Pro Model Card (June 2025 GA): Video-MME (audio+visual+subtitles) 86.9%
gemini-2.5-pro,aider_polyglot_diff,82.2,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-5-Pro-Model-Card.pdf,"{""effort"":""default"",""harness"":""AI Studio API"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""diff-fenced"",""sampling"":""pass@1 (single attempt)"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.5 Pro Model Card (June 2025 GA): Aider Polyglot diff-fenced 82.2% (avg of 3 trials)
gemini-2.5-pro,vibe_eval,67.2,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-5-Pro-Model-Card.pdf,"{""effort"":""default"",""harness"":""AI Studio API"",""judge"":""Gemini"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (single attempt)"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.5 Pro Model Card (June 2025 GA): Vibe-Eval (Reka) 67.2% (Gemini-as-judge)
gemini-2.5-pro,global_mmlu_lite,89.2,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-5-Pro-Model-Card.pdf,"{""effort"":""default"",""harness"":""AI Studio API"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (single attempt)"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,"Gemini 2.5 Pro Model Card (GA, June 27 2025): Global MMLU Lite 89.2%."
gpt-4.1-mini,aime_2024,49.6,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1-mini=49.6.
gpt-4.1-nano,aime_2024,29.4,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1-nano=29.4.
gpt-4o,aime_2024,13.1,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o=13.1.
gpt-4o-mini,aime_2024,8.6,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o-mini=8.6.
o1-high,aime_2024,74.3,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o1-high=74.3.
gpt-4o,gpqa_diamond,46.0,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o=46.0.
gpt-4o-mini,gpqa_diamond,40.2,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o-mini=40.2.
o1-high,gpqa_diamond,75.7,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o1-high=75.7.
gpt-4o,mmlu,85.7,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o=85.7.
gpt-4o-mini,mmlu,82.0,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o-mini=82.0.
o1-high,mmlu,91.8,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o1-high=91.8.
gpt-4.1,mmmlu,87.3,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1=87.3.
gpt-4.1-mini,mmmlu,78.5,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1-mini=78.5.
gpt-4.1-nano,mmmlu,66.9,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1-nano=66.9.
gpt-4o,mmmlu,81.4,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o=81.4.
gpt-4o-mini,mmmlu,70.5,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o-mini=70.5.
o1-high,mmmlu,87.7,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o1-high=87.7.
o3-mini-high,mmmlu,80.7,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o3-mini-high=80.7.
gpt-4.5,mmmlu,85.1,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.5=85.1.
gpt-4o,swe_bench_verified,30.7,https://openai.com/index/introducing-gpt-4-5/,"{""effort"":""n/a"",""harness"":""official (GPT-4.5 blog)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,OpenAI GPT-4.5 blog: gpt-4o=30.7. (GPT-4.1 blog 477-subset value 33.2 dropped — was different N).
gpt-4o-mini,swe_bench_verified,8.7,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,dropped,"GPT-4.1 blog value 8.7 was 477/500 subset, not full SWE-bench Verified. No verified full-set value available; dropped. DROPPED: source uses 477/500 subset (omit 23 problems), not full SWE-bench Verified."
o1-high,swe_bench_verified,48.9,https://www.anthropic.com/news/claude-3-7-sonnet,"{""effort"":""high"",""harness"":""Anthropic simple scaffold"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash+editor)""}",False,official_blog,verified,Anthropic 3.7 Sonnet blog (third-party eval): o1 SWE-bench Verified=48.9%. Note: diff scaffold from OpenAI. Overwrite dropped cell (41.0 from gpt-4-1 blog was dropped).
gpt-4o,aider_polyglot_diff,18.2,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o=18.2.
gpt-4o-mini,aider_polyglot_diff,2.7,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o-mini=2.7.
o1-high,aider_polyglot_diff,61.7,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o1-high=61.7.
o3-mini-high,aider_polyglot_diff,60.4,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o3-mini-high=60.4.
gpt-4.5,aider_polyglot_diff,44.9,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.5=44.9.
gpt-4.1,aider_polyglot_whole,51.6,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1=51.6.
gpt-4.1-mini,aider_polyglot_whole,34.7,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1-mini=34.7.
gpt-4.1-nano,aider_polyglot_whole,9.8,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1-nano=9.8.
gpt-4o,aider_polyglot_whole,30.7,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o=30.7.
gpt-4o-mini,aider_polyglot_whole,3.6,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o-mini=3.6.
o1-high,aider_polyglot_whole,64.6,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o1-high=64.6.
o3-mini-high,aider_polyglot_whole,66.7,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o3-mini-high=66.7.
gpt-4o,internal_api_if_hard,29.2,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o=29.2.
gpt-4o-mini,internal_api_if_hard,27.2,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o-mini=27.2.
o1-high,internal_api_if_hard,51.3,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o1-high=51.3.
o3-mini-high,internal_api_if_hard,50.0,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o3-mini-high=50.0.
gpt-4.5,internal_api_if_hard,54.0,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.5=54.0.
gpt-4.1,multichallenge,38.3,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1=38.3.
gpt-4.1-mini,multichallenge,35.8,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1-mini=35.8.
gpt-4.1-nano,multichallenge,15.0,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1-nano=15.0.
gpt-4o,multichallenge,27.8,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o=27.8.
gpt-4o-mini,multichallenge,20.3,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o-mini=20.3.
o1-high,multichallenge,44.9,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o1-high=44.9.
o3-mini-high,multichallenge,39.9,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o3-mini-high=39.9.
gpt-4.5,multichallenge,43.8,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.5=43.8.
gpt-4o,multichallenge_o3mini_grader,39.9,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o=39.9.
gpt-4o-mini,multichallenge_o3mini_grader,25.6,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o-mini=25.6.
o1-high,multichallenge_o3mini_grader,52.9,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o1-high=52.9.
o3-mini-high,multichallenge_o3mini_grader,50.2,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o3-mini-high=50.2.
gpt-4.5,multichallenge_o3mini_grader,50.1,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.5=50.1.
gpt-4o,collie,50.2,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o=50.2.
gpt-4o-mini,collie,52.7,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o-mini=52.7.
o1-high,collie,95.3,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o1-high=95.3.
o3-mini-high,collie,98.7,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o3-mini-high=98.7.
gpt-4.5,collie,72.3,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.5=72.3.
gpt-4o,ifeval,81.0,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o=81.0.
gpt-4o-mini,ifeval,78.4,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o-mini=78.4.
o1-high,ifeval,92.2,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o1-high=92.2.
gpt-4.1,multi_if,70.8,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1=70.8.
gpt-4.1-mini,multi_if,67.0,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1-mini=67.0.
gpt-4.1-nano,multi_if,57.2,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1-nano=57.2.
gpt-4o,multi_if,60.9,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o=60.9.
gpt-4o-mini,multi_if,57.9,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o-mini=57.9.
o1-high,multi_if,77.9,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o1-high=77.9.
o3-mini-high,multi_if,79.5,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o3-mini-high=79.5.
gpt-4.5,multi_if,70.8,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.5=70.8.
gpt-4.1,mrcr_v2_2needle_128k,57.2,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1=57.2.
gpt-4.1-mini,mrcr_v2_2needle_128k,47.2,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1-mini=47.2.
gpt-4.1-nano,mrcr_v2_2needle_128k,36.6,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1-nano=36.6.
gpt-4o,mrcr_v2_2needle_128k,31.9,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o=31.9.
gpt-4o-mini,mrcr_v2_2needle_128k,24.5,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o-mini=24.5.
o1-high,mrcr_v2_2needle_128k,22.1,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o1-high=22.1.
o3-mini-high,mrcr_v2_2needle_128k,18.7,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o3-mini-high=18.7.
gpt-4.5,mrcr_v2_2needle_128k,38.5,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.5=38.5.
gpt-4.1,mrcr_v2_2needle_1m,46.3,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1=46.3.
gpt-4.1-mini,mrcr_v2_2needle_1m,33.3,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1-mini=33.3.
gpt-4.1-nano,mrcr_v2_2needle_1m,12.0,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1-nano=12.0.
gpt-4o,graphwalks_bfs_0k_128k,41.7,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o=41.7.
gpt-4o-mini,graphwalks_bfs_0k_128k,29.0,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o-mini=29.0.
o1-high,graphwalks_bfs_0k_128k,62.0,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o1-high=62.0.
o3-mini-high,graphwalks_bfs_0k_128k,51.0,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o3-mini-high=51.0.
gpt-4.5,graphwalks_bfs_0k_128k,72.3,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.5=72.3.
gpt-4.1,graphwalks_bfs_128k_plus,19.0,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1=19.0.
gpt-4.1-mini,graphwalks_bfs_128k_plus,15.0,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1-mini=15.0.
gpt-4.1-nano,graphwalks_bfs_128k_plus,2.9,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1-nano=2.9.
gpt-4o,graphwalks_parents_0k_128k,35.4,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o=35.4.
gpt-4o-mini,graphwalks_parents_0k_128k,12.6,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o-mini=12.6.
o1-high,graphwalks_parents_0k_128k,50.9,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o1-high=50.9.
o3-mini-high,graphwalks_parents_0k_128k,58.3,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o3-mini-high=58.3.
gpt-4.5,graphwalks_parents_0k_128k,72.6,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.5=72.6.
gpt-4.1,graphwalks_parents_128k_plus,25.0,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1=25.0.
gpt-4.1-mini,graphwalks_parents_128k_plus,11.0,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1-mini=11.0.
gpt-4.1-nano,graphwalks_parents_128k_plus,5.6,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1-nano=5.6.
gpt-4o,mmmu,68.7,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o=68.7.
gpt-4o-mini,mmmu,56.3,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o-mini=56.3.
o1-high,mmmu,77.6,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o1-high=77.6.
gpt-4.5,mmmu,75.2,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.5=75.2.
gpt-4.1,mathvista,72.2,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1=72.2.
gpt-4.1-mini,mathvista,73.1,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1-mini=73.1.
gpt-4.1-nano,mathvista,56.2,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1-nano=56.2.
gpt-4o,mathvista,61.4,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o=61.4.
gpt-4o-mini,mathvista,56.5,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o-mini=56.5.
o1-high,mathvista,71.8,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o1-high=71.8.
gpt-4.5,mathvista,72.3,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.5=72.3.
gpt-4o,charxiv_reasoning,52.7,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o=52.7.
gpt-4o-mini,charxiv_reasoning,36.8,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o-mini=36.8.
o1-high,charxiv_reasoning,55.1,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o1-high=55.1.
gpt-4.5,charxiv_reasoning,55.4,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.5=55.4.
gpt-4.1,charxiv_descriptive,87.9,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1=87.9.
gpt-4.1-mini,charxiv_descriptive,88.4,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1-mini=88.4.
gpt-4.1-nano,charxiv_descriptive,73.9,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1-nano=73.9.
gpt-4o,charxiv_descriptive,85.3,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o=85.3.
gpt-4o-mini,charxiv_descriptive,76.6,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o-mini=76.6.
o1-high,charxiv_descriptive,88.9,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o1-high=88.9.
gpt-4.5,charxiv_descriptive,90.0,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.5=90.0.
gpt-4.1,complexfuncbench,65.5,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (function calling)""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1=65.5.
gpt-4.1-mini,complexfuncbench,49.3,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (function calling)""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1-mini=49.3.
gpt-4.1-nano,complexfuncbench,5.7,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (function calling)""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1-nano=5.7.
gpt-4o,complexfuncbench,66.5,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (function calling)""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o=66.5.
gpt-4o-mini,complexfuncbench,38.6,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (function calling)""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o-mini=38.6.
o1-high,complexfuncbench,47.6,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (function calling)""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o1-high=47.6.
o3-mini-high,complexfuncbench,17.6,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (function calling)""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o3-mini-high=17.6.
gpt-4.5,complexfuncbench,63.0,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (function calling)""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.5=63.0.
gpt-4.1,tau_bench_airline,49.4,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1=49.4.
gpt-4.1-mini,tau_bench_airline,36.0,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1-mini=36.0.
gpt-4.1-nano,tau_bench_airline,14.0,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1-nano=14.0.
gpt-4o,tau_bench_airline,42.8,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o=42.8.
gpt-4o-mini,tau_bench_airline,22.0,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o-mini=22.0.
o1-high,tau_bench_airline,50.0,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o1-high=50.0.
o3-mini-high,tau_bench_airline,32.4,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o3-mini-high=32.4.
gpt-4.5,tau_bench_airline,50.0,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.5=50.0.
gpt-4.1,tau_bench_retail,68.0,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1=68.0.
gpt-4.1-mini,tau_bench_retail,55.8,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1-mini=55.8.
gpt-4.1-nano,tau_bench_retail,22.6,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.1-nano=22.6.
gpt-4o,tau_bench_retail,60.3,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o=60.3.
gpt-4o-mini,tau_bench_retail,44.0,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4o-mini=44.0.
o1-high,tau_bench_retail,70.8,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o1-high=70.8.
o3-mini-high,tau_bench_retail,57.6,https://openai.com/index/gpt-4-1/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-4.1 blog: o3-mini-high=57.6.
gpt-4.5,tau_bench_retail,68.4,https://openai.com/index/gpt-4-1/,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,OpenAI GPT-4.1 blog: gpt-4.5=68.4.
gemini-1.5-flash,mmlu_pro,67.3,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): MMLU-Pro 67.3%
gemini-1.5-pro,mmlu_pro,75.8,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): MMLU-Pro 75.8%
gemini-2.0-flash-lite,mmlu_pro,71.6,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): MMLU-Pro 71.6%
gemini-1.5-flash,livecodebench,30.7,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): LiveCodeBench v5 (UI 10/01/2024-02/01/2025) 30.7%
gemini-1.5-pro,livecodebench,34.2,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): LiveCodeBench v5 (UI 10/01/2024-02/01/2025) 34.2%
gemini-2.0-flash-lite,livecodebench,28.9,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): LiveCodeBench v5 (UI 10/01/2024-02/01/2025) 28.9%
gemini-1.5-flash,bird_sql,45.6,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): Bird-SQL Dev (NL->SQL) 45.6%
gemini-1.5-pro,bird_sql,54.4,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): Bird-SQL Dev (NL->SQL) 54.4%
gemini-2.0-flash-lite,bird_sql,57.4,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): Bird-SQL Dev (NL->SQL) 57.4%
gemini-2.0-flash,bird_sql,58.7,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): Bird-SQL Dev (NL->SQL) 58.7%
gemini-1.5-flash,gpqa_diamond,51.0,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): GPQA Diamond 51.0%
gemini-1.5-pro,gpqa_diamond,59.1,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): GPQA Diamond 59.1%
gemini-2.0-flash-lite,gpqa_diamond,51.5,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): GPQA Diamond 51.5%
gemini-1.5-flash,simpleqa,8.6,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): SimpleQA (no search) 8.6%
gemini-1.5-pro,simpleqa,24.9,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): SimpleQA (no search) 24.9%
gemini-2.0-flash-lite,simpleqa,21.7,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): SimpleQA (no search) 21.7%
gemini-1.5-flash,facts_grounding,82.9,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): FACTS Grounding 82.9%
gemini-1.5-pro,facts_grounding,80.0,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): FACTS Grounding 80.0%
gemini-2.0-flash-lite,facts_grounding,83.6,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): FACTS Grounding 83.6%
gemini-2.0-flash,facts_grounding,84.6,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): FACTS Grounding 84.6%
gemini-1.5-flash,global_mmlu_lite,73.7,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): Global MMLU Lite 73.7%
gemini-1.5-pro,global_mmlu_lite,80.8,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): Global MMLU Lite 80.8%
gemini-2.0-flash-lite,global_mmlu_lite,78.2,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): Global MMLU Lite 78.2%
gemini-2.0-flash,global_mmlu_lite,83.4,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): Global MMLU Lite 83.4%
gemini-1.5-flash,math,77.9,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): MATH (full) 77.9%
gemini-1.5-pro,math,86.5,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): MATH (full) 86.5%
gemini-2.0-flash-lite,math,86.8,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): MATH (full) 86.8%
gemini-2.0-flash,math,90.9,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): MATH (full) 90.9%
gemini-1.5-flash,hiddenmath,47.2,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): HiddenMath 47.2%
gemini-1.5-pro,hiddenmath,52.0,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): HiddenMath 52.0%
gemini-2.0-flash-lite,hiddenmath,55.3,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): HiddenMath 55.3%
gemini-2.0-flash,hiddenmath,63.5,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): HiddenMath 63.5%
gemini-1.5-flash,mrcr_v1,71.9,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): MRCR v1 1M 71.9%
gemini-1.5-pro,mrcr_v1,82.6,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): MRCR v1 1M 82.6%
gemini-2.0-flash-lite,mrcr_v1,58.0,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): MRCR v1 1M 58.0%
gemini-2.0-flash,mrcr_v1,70.5,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): MRCR v1 1M 70.5%
gemini-1.5-flash,mmmu,62.3,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): MMMU 62.3%
gemini-1.5-pro,mmmu,65.9,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): MMMU 65.9%
gemini-2.0-flash-lite,mmmu,68.0,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): MMMU 68.0%
gemini-1.5-flash,covost2,37.4,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): CoVoST2 21-lang BLEU 37.4 BLEU
gemini-1.5-pro,covost2,40.1,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): CoVoST2 21-lang BLEU 40.1 BLEU
gemini-2.0-flash-lite,covost2,38.4,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): CoVoST2 21-lang BLEU 38.4 BLEU
gemini-2.0-flash,covost2,39.0,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): CoVoST2 21-lang BLEU 39.0 BLEU
gemini-1.5-flash,egoschema,66.8,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): EgoSchema test 66.8%
gemini-1.5-pro,egoschema,71.2,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): EgoSchema test 71.2%
gemini-2.0-flash-lite,egoschema,67.2,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): EgoSchema test 67.2%
gemini-2.0-flash,egoschema,71.1,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-0-Flash-Model-Card.pdf,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.0 Flash Model Card (April 15 2025): EgoSchema test 71.1%
grok-3-beta,mmmu,73.2,https://x.ai/news/grok-3,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,xAI Grok 3 blog (cross-model non-reasoning table): grok-3-beta=73.2.
grok-3-beta,loft_128k,83.3,https://x.ai/news/grok-3,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,xAI Grok 3 blog (cross-model non-reasoning table): grok-3-beta=83.3.
grok-3-beta,egoschema,74.5,https://x.ai/news/grok-3,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,xAI Grok 3 blog (cross-model non-reasoning table): grok-3-beta=74.5.
claude-sonnet-4.6,cybergym,65.2,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""cybergym"",""sampling"":""pass@1"",""tools"":""cyber execution environment""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 3.2.3.A, page 33: Claude Sonnet 4.6; CyberGym vulnerability discovery [targeted]=65.2%. Source setting: effort=source does not state; tools=cyber execution environment; sampling=pass@1; harness=cybergym."
claude-sonnet-4.6,vending_bench_2,7204.14,https://anthropic.com/claude-sonnet-4-6-system-card,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""max effort"",""prompt_style"":""default"",""sampling"":""final balance after 1-year simulation"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,"Claude Sonnet 4.6 System Card section 2.13: Vending-Bench 2 final balance $7,204.14 with Max effort (vs Opus 4.6 SOTA $8017.59). High-effort run scored $6,625.10 (kept as candidate setting note). Mean cost per run $265.03."
grok-4.1,eq_bench3,1585.0,https://x.ai/news/grok-4-1,"{""effort"":""default"",""harness"":""official"",""judge"":""LLM judge"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,xAI Grok 4.1 blog: grok-4.1=1585 (Elo Normalized).
grok-4.1,creative_writing_v3,1708.6,https://x.ai/news/grok-4-1,"{""effort"":""default"",""harness"":""official"",""judge"":""LLM judge"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,xAI Grok 4.1 blog: grok-4.1=1708.6 (Elo Normalized).
qwen3-235b,mmlu_redux,92.7,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report (arxiv:2505.09388) Table 11 Thinking column: MMLU-Redux 92.7
qwen3-235b,bfcl_v3,70.8,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report (arxiv:2505.09388) Table 11 Thinking column: BFCL v3 70.8
qwen3-235b,multi_if,71.9,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report (arxiv:2505.09388) Table 11 Thinking column: Multi-IF 71.9
qwen3-235b,mt_aime_2024,80.8,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report (arxiv:2505.09388) Table 11 Thinking column: MT-AIME2024 80.8
qwen3-235b,mmmlu,84.3,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report (arxiv:2505.09388) Table 11 Thinking column: MMMLU 14 languages 84.3
qwen3-235b,livebench,77.1,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report (arxiv:2505.09388) Table 11 Thinking column: LiveBench 2024-11-25 77.1
deepseek-v4-pro,chinese_simpleqa,84.4,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-pro (Max mode) = 84.4.
deepseek-v4-pro,mrcr_v1,83.5,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-pro (Max mode) = 83.5.
deepseek-v4-pro,corpusqa_1m,62.0,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-pro (Max mode) = 62.0.
deepseek-v4-flash,chinese_simpleqa,78.9,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-flash (Max mode) = 78.9.
deepseek-v4-flash,mrcr_v1,78.7,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-flash (Max mode) = 78.7.
deepseek-v4-flash,corpusqa_1m,60.5,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4 model card: deepseek-v4-flash (Max mode) = 60.5.
gpt-5.4,mmlu_pro,87.5,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): gpt-5.4 (per their own canonical config) = 87.5.
kimi-k2.6,mmlu_pro,87.1,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): kimi-k2.6 (per their own canonical config) = 87.1.
glm-5.1,mmlu_pro,86.0,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): glm-5.1 (per their own canonical config) = 86.0.
claude-opus-4.6,chinese_simpleqa,76.4,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): claude-opus-4.6 (per their own canonical config) = 76.4.
gpt-5.4,chinese_simpleqa,76.8,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): gpt-5.4 (per their own canonical config) = 76.8.
gemini-3.1-pro,chinese_simpleqa,85.9,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): gemini-3.1-pro (per their own canonical config) = 85.9.
kimi-k2.6,chinese_simpleqa,75.9,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): kimi-k2.6 (per their own canonical config) = 75.9.
glm-5.1,chinese_simpleqa,75.0,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): glm-5.1 (per their own canonical config) = 75.0.
gpt-5.4,codeforces_rating,3168.0,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): gpt-5.4 (per their own canonical config) = 3168.
gemini-3.1-pro,imo_answerbench,81.0,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): gemini-3.1-pro (per their own canonical config) = 81.0.
claude-opus-4.6,matharena_apex_2025,34.5,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): claude-opus-4.6 (per their own canonical config) = 34.5.
gpt-5.4,matharena_apex_2025,54.1,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): gpt-5.4 (per their own canonical config) = 54.1.
kimi-k2.6,matharena_apex_2025,24.0,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): kimi-k2.6 (per their own canonical config) = 24.0.
glm-5.1,matharena_apex_2025,11.5,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): glm-5.1 (per their own canonical config) = 11.5.
claude-opus-4.6,mrcr_v1,92.9,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): claude-opus-4.6 (per their own canonical config) = 92.9.
gemini-3.1-pro,mrcr_v1,76.3,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): gemini-3.1-pro (per their own canonical config) = 76.3.
claude-opus-4.6,corpusqa_1m,71.7,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): claude-opus-4.6 (per their own canonical config) = 71.7.
gemini-3.1-pro,corpusqa_1m,53.8,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): gemini-3.1-pro (per their own canonical config) = 53.8.
glm-5.1,swe_bench_multilingual,73.3,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): glm-5.1 (per their own canonical config) = 73.3.
gpt-5.4,gdpval_aa_elo,1674.0,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): gpt-5.4 (per their own canonical config) = 1674.
gemini-3.1-pro,gdpval_aa_elo,1314.0,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): gemini-3.1-pro (per their own canonical config) = 1314.
kimi-k2.6,gdpval_aa_elo,1482.0,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): kimi-k2.6 (per their own canonical config) = 1482.
glm-5.1,gdpval_aa_elo,1535.0,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek V4-Pro model card (third-party): glm-5.1 (per their own canonical config) = 1535.
qwen3-32b,mmlu_redux,90.9,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: MMLU-Redux 90.9
qwen3-32b,bfcl_v3,70.3,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: BFCL v3 70.3
qwen3-32b,multi_if,73.0,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: Multi-IF 73.0
qwen3-32b,mt_aime_2024,75.0,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: MT-AIME2024 75.0
qwen3-32b,mmmlu,80.6,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: MMMLU 14 langs 80.6
qwen3-32b,livebench,74.9,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: LiveBench 2024-11-25 74.9
qwen3-30b-a3b,mmlu_redux,89.5,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: MMLU-Redux 89.5
qwen3-30b-a3b,bfcl_v3,69.1,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: BFCL v3 69.1
qwen3-30b-a3b,multi_if,72.2,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: Multi-IF 72.2
qwen3-30b-a3b,mt_aime_2024,73.9,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: MT-AIME2024 73.9
qwen3-30b-a3b,mmmlu,78.4,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: MMMLU 14 langs 78.4
qwen3-30b-a3b,livebench,74.3,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: LiveBench 2024-11-25 74.3
qwen3-14b,mmlu_redux,88.6,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: MMLU-Redux 88.6
qwen3-14b,bfcl_v3,70.4,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: BFCL v3 70.4
qwen3-14b,multi_if,74.8,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: Multi-IF 74.8
qwen3-14b,mt_aime_2024,73.3,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: MT-AIME2024 73.3
qwen3-14b,mmmlu,77.9,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: MMMLU 14 langs 77.9
qwen3-14b,livebench,71.3,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: LiveBench 2024-11-25 71.3
qwen3-14b,ifeval,85.4,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: IFEval strict 85.4
qwen3-8b,mmlu_redux,87.5,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: MMLU-Redux 87.5
qwen3-8b,bfcl_v3,68.1,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: BFCL v3 68.1
qwen3-8b,multi_if,71.2,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: Multi-IF 71.2
qwen3-8b,mt_aime_2024,65.4,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: MT-AIME2024 65.4
qwen3-8b,mmmlu,74.4,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: MMMLU 14 langs 74.4
qwen3-8b,livebench,67.1,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: LiveBench 2024-11-25 67.1
qwen3-4b,mmlu_redux,83.7,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: MMLU-Redux 83.7
qwen3-4b,bfcl_v3,65.9,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: BFCL v3 65.9
qwen3-4b,multi_if,66.3,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: Multi-IF 66.3
qwen3-4b,mt_aime_2024,60.7,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: MT-AIME2024 60.7
qwen3-4b,mmmlu,69.8,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: MMMLU 14 langs 69.8
qwen3-4b,livebench,63.6,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: LiveBench 2024-11-25 63.6
qwen3-1.7b,mmlu_redux,73.9,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: MMLU-Redux 73.9
qwen3-1.7b,bfcl_v3,56.6,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: BFCL v3 56.6
qwen3-1.7b,multi_if,51.2,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: Multi-IF 51.2
qwen3-1.7b,mt_aime_2024,36.1,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: MT-AIME2024 36.1
qwen3-1.7b,mmmlu,59.1,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: MMMLU 14 langs 59.1
qwen3-0.6b,mmlu_redux,55.6,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: MMLU-Redux 55.6
qwen3-0.6b,bfcl_v3,46.4,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: BFCL v3 46.4
qwen3-0.6b,multi_if,36.1,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: Multi-IF 36.1
qwen3-0.6b,mt_aime_2024,7.8,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: MT-AIME2024 7.8
qwen3-0.6b,mmmlu,43.1,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Qwen3 tech report Thinking: MMMLU 14 langs 43.1
claude-sonnet-4.5,hmmt_feb_2025,79.2,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: claude-sonnet-4.5=79.2.
gpt-5,hmmt_feb_2025,88.3,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: gpt-5=88.3.
gemini-3-pro,hmmt_feb_2025,97.5,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: gemini-3-pro=97.5.
kimi-k2-thinking,hmmt_feb_2025,89.4,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: kimi-k2-thinking=89.4.
deepseek-v3.2,hmmt_feb_2025,92.5,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: deepseek-v3.2=92.5.
claude-sonnet-4.5,hmmt_nov_2025,81.7,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: claude-sonnet-4.5=81.7.
gpt-5,imo_answerbench,76.0,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: gpt-5=76.0.
gemini-3-pro,imo_answerbench,83.3,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: gemini-3-pro=83.3.
deepseek-v3.2,imo_answerbench,78.3,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: deepseek-v3.2=78.3.
claude-sonnet-4.5,swe_bench_multilingual,68.0,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: claude-sonnet-4.5=68.0.
gpt-5,swe_bench_multilingual,55.3,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: gpt-5=55.3.
kimi-k2-thinking,swe_bench_multilingual,61.1,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: kimi-k2-thinking=61.1.
minimax-m2,swe_bench_multilingual,56.5,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: minimax-m2=56.5.
deepseek-v3.2,swe_bench_multilingual,70.2,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic (bash + edit)""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: deepseek-v3.2=70.2.
deepseek-v3.2,tau2_bench_airline,63.8,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: deepseek-v3.2=63.8.
deepseek-v3.2,tau2_bench_retail,81.1,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: deepseek-v3.2=81.1.
deepseek-v3.2,tau2_bench_telecom,96.2,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,tech_report,verified,DeepSeek V3.2 tech report Table 2: deepseek-v3.2=96.2.
kimi-k2,mmlu_redux,92.7,https://arxiv.org/abs/2507.20534,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg of 4-64 trials per benchmark)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,tech_report,verified,Kimi K2 tech report (arxiv:2507.20534) Table 3 Kimi-K2-Instruct column: MMLU-Redux EM 92.7
kimi-k2,swe_bench_multilingual,47.3,https://arxiv.org/abs/2507.20534,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg of 4-64 trials per benchmark)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,tech_report,verified,Kimi K2 tech report (arxiv:2507.20534) Table 3 Kimi-K2-Instruct column: SWE-bench Multilingual Pass@1 47.3
kimi-k2,tau2_bench_retail,70.6,https://arxiv.org/abs/2507.20534,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg of 4-64 trials per benchmark)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,tech_report,verified,Kimi K2 tech report (arxiv:2507.20534) Table 3 Kimi-K2-Instruct column: Tau2 retail Avg@4 70.6
kimi-k2,tau2_bench_airline,56.5,https://arxiv.org/abs/2507.20534,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg of 4-64 trials per benchmark)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,tech_report,verified,Kimi K2 tech report (arxiv:2507.20534) Table 3 Kimi-K2-Instruct column: Tau2 airline Avg@4 56.5
kimi-k2,tau2_bench_telecom,65.8,https://arxiv.org/abs/2507.20534,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg of 4-64 trials per benchmark)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,tech_report,verified,Kimi K2 tech report (arxiv:2507.20534) Table 3 Kimi-K2-Instruct column: Tau2 telecom Avg@4 65.8
kimi-k2,supergpqa,57.2,https://arxiv.org/abs/2507.20534,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg of 4-64 trials per benchmark)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,tech_report,verified,Kimi K2 tech report (arxiv:2507.20534) Table 3 Kimi-K2-Instruct column: SuperGPQA Acc 57.2
kimi-k2,livebench,76.4,https://arxiv.org/abs/2507.20534,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg of 4-64 trials per benchmark)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,tech_report,verified,Kimi K2 tech report (arxiv:2507.20534) Table 3 Kimi-K2-Instruct column: Livebench Pass@1 76.4
kimi-k2-thinking,imo_answerbench,78.6,https://moonshotai.github.io/Kimi-K2/thinking.html,"{""effort"":""default (96k thinking-token budget for HLE/AIME/HMMT/GPQA; 128k for IMO/LCB/OJ)"",""harness"":""official"",""judge"":""rule-based (HLE uses o3-mini)"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 for AIME/HMMT no-tools)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,official_blog,verified,Kimi K2 Thinking blog cross-model table: IMO-AnswerBench no-tools avg@8 78.6
kimi-k2-thinking,mmlu_redux,94.4,https://moonshotai.github.io/Kimi-K2/thinking.html,"{""effort"":""default (96k thinking-token budget for HLE/AIME/HMMT/GPQA; 128k for IMO/LCB/OJ)"",""harness"":""official"",""judge"":""rule-based (HLE uses o3-mini)"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 for AIME/HMMT no-tools)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,official_blog,verified,Kimi K2 Thinking blog cross-model table: MMLU-Redux no-tools 94.4
kimi-k2-thinking,multi_swe_bench,41.9,https://moonshotai.github.io/Kimi-K2/thinking.html,"{""effort"":""default (96k thinking-token budget for HLE/AIME/HMMT/GPQA; 128k for IMO/LCB/OJ)"",""harness"":""official"",""judge"":""rule-based (HLE uses o3-mini)"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 for AIME/HMMT no-tools)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,official_blog,verified,Kimi K2 Thinking blog cross-model table: Multi-SWE-bench w/ tools 41.9
kimi-k2-thinking,scicode,44.8,https://moonshotai.github.io/Kimi-K2/thinking.html,"{""effort"":""default (96k thinking-token budget for HLE/AIME/HMMT/GPQA; 128k for IMO/LCB/OJ)"",""harness"":""official"",""judge"":""rule-based (HLE uses o3-mini)"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 for AIME/HMMT no-tools)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,official_blog,verified,Kimi K2 Thinking blog cross-model table: SciCode no-tools 44.8
kimi-k2-thinking,ojbench,48.7,https://moonshotai.github.io/Kimi-K2/thinking.html,"{""effort"":""default (96k thinking-token budget for HLE/AIME/HMMT/GPQA; 128k for IMO/LCB/OJ)"",""harness"":""official"",""judge"":""rule-based (HLE uses o3-mini)"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 for AIME/HMMT no-tools)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,official_blog,verified,Kimi K2 Thinking blog cross-model table: OJ-Bench cpp no-tools 48.7
kimi-k2-thinking,browsecomp_zh,62.3,https://moonshotai.github.io/Kimi-K2/thinking.html,"{""effort"":""default (96k thinking-token budget for HLE/AIME/HMMT/GPQA; 128k for IMO/LCB/OJ)"",""harness"":""official"",""judge"":""rule-based (HLE uses o3-mini)"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 for AIME/HMMT no-tools)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,official_blog,verified,Kimi K2 Thinking blog cross-model table: BrowseComp-ZH w/ tools 62.3
kimi-k2-thinking,frames,87.0,https://moonshotai.github.io/Kimi-K2/thinking.html,"{""effort"":""default (96k thinking-token budget for HLE/AIME/HMMT/GPQA; 128k for IMO/LCB/OJ)"",""harness"":""official"",""judge"":""rule-based (HLE uses o3-mini)"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 for AIME/HMMT no-tools)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,official_blog,verified,Kimi K2 Thinking blog cross-model table: Frames w/ tools 87.0
kimi-k2.5,mathvista_mini,90.1,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default (96k for reasoning, 64k for vision)"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 AIME/HMMT, avg@8 GPQA, avg@3 vision, avg@4 Seal/WideSearch)"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,model_card,verified,Kimi K2.5 model card (HF moonshotai/Kimi-K2.5) Evaluation table: MathVista mini 90.1
kimi-k2.5,ocrbench,92.3,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default (96k for reasoning, 64k for vision)"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 AIME/HMMT, avg@8 GPQA, avg@3 vision, avg@4 Seal/WideSearch)"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,model_card,verified,Kimi K2.5 model card (HF moonshotai/Kimi-K2.5) Evaluation table: OCRBench 92.3
kimi-k2.5,omnidocbench,0.112,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default (96k for reasoning, 64k for vision)"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 AIME/HMMT, avg@8 GPQA, avg@3 vision, avg@4 Seal/WideSearch)"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,model_card,verified,Kimi K2.5 model card (HF moonshotai/Kimi-K2.5) Evaluation table: OmniDocBench 1.5 (1 - normalized Levenshtein) ×100 = 88.8 [Converted 88.8% (1-LED higher-better) → 0.112 (edit distance lower-better) for cross-source consistency]
kimi-k2.5,longbench_v2,61.0,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default (96k for reasoning, 64k for vision)"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 AIME/HMMT, avg@8 GPQA, avg@3 vision, avg@4 Seal/WideSearch)"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,model_card,verified,Kimi K2.5 model card (HF moonshotai/Kimi-K2.5) Evaluation table: LongBench v2 61.0
kimi-k2.5,paperbench,63.5,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default (96k for reasoning, 64k for vision)"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 AIME/HMMT, avg@8 GPQA, avg@3 vision, avg@4 Seal/WideSearch)"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,model_card,verified,Kimi K2.5 model card (HF moonshotai/Kimi-K2.5) Evaluation table: PaperBench 63.5
kimi-k2.5,cybergym,41.3,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default (96k for reasoning, 64k for vision)"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 AIME/HMMT, avg@8 GPQA, avg@3 vision, avg@4 Seal/WideSearch)"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,model_card,verified,Kimi K2.5 model card (HF moonshotai/Kimi-K2.5) Evaluation table: CyberGym 41.3
kimi-k2.5,ojbench,57.4,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default (96k for reasoning, 64k for vision)"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 AIME/HMMT, avg@8 GPQA, avg@3 vision, avg@4 Seal/WideSearch)"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,model_card,verified,Kimi K2.5 model card (HF moonshotai/Kimi-K2.5) Evaluation table: OJBench cpp 57.4
kimi-k2.5,video_mme,87.4,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default (96k for reasoning, 64k for vision)"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 AIME/HMMT, avg@8 GPQA, avg@3 vision, avg@4 Seal/WideSearch)"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,model_card,verified,Kimi K2.5 model card (HF moonshotai/Kimi-K2.5) Evaluation table: VideoMME 87.4
kimi-k2.6,babyvision,39.8,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""max (98304 generation tokens)"",""harness"":""official (in-house SWE-agent-derived for SWE-Bench)"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@10 coding, avg@3 vision)"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,model_card,verified,Kimi K2.6 model card (HF moonshotai/Kimi-K2.6) Evaluation table: BabyVision 39.8
kimi-k2.6,claw_eval_pass3,62.3,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""max (98304 generation tokens)"",""harness"":""official (in-house SWE-agent-derived for SWE-Bench)"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@10 coding, avg@3 vision)"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,model_card,verified,Kimi K2.6 model card (HF moonshotai/Kimi-K2.6) Evaluation table: Claw Eval pass^3 62.3
kimi-k2.6,mcpmark,55.9,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""max (98304 generation tokens)"",""harness"":""official (in-house SWE-agent-derived for SWE-Bench)"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg@10 coding, avg@3 vision)"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,model_card,verified,Kimi K2.6 model card (HF moonshotai/Kimi-K2.6) Evaluation table: MCPMark 55.9
deepseek-v3.2-speciale,imo_answerbench,84.5,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,tech_report,verified,DS V3.2 tech report Table 3: V3.2-Speciale=84.5.
deepseek-r1,mmlu_redux,92.9,https://huggingface.co/deepseek-ai/DeepSeek-R1-0528,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek R1-0528 model card: deepseek-r1=92.9.
deepseek-r1-0528,mmlu_redux,93.4,https://huggingface.co/deepseek-ai/DeepSeek-R1-0528,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek R1-0528 model card: deepseek-r1-0528=93.4.
deepseek-r1,frames,82.5,https://huggingface.co/deepseek-ai/DeepSeek-R1-0528,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek R1-0528 model card: deepseek-r1=82.5.
deepseek-r1-0528,frames,83.0,https://huggingface.co/deepseek-ai/DeepSeek-R1-0528,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek R1-0528 model card: deepseek-r1-0528=83.0.
deepseek-r1,aider_polyglot_diff,53.3,https://huggingface.co/deepseek-ai/DeepSeek-R1-0528,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,model_card,verified,DeepSeek R1-0528 model card: deepseek-r1=53.3.
deepseek-r1-0528,aider_polyglot_diff,71.6,https://huggingface.co/deepseek-ai/DeepSeek-R1-0528,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,model_card,verified,DeepSeek R1-0528 model card: deepseek-r1-0528=71.6.
deepseek-r1,cnmo_2024,78.8,https://huggingface.co/deepseek-ai/DeepSeek-R1-0528,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek R1-0528 model card: deepseek-r1=78.8.
deepseek-r1-0528,cnmo_2024,86.9,https://huggingface.co/deepseek-ai/DeepSeek-R1-0528,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,DeepSeek R1-0528 model card: deepseek-r1-0528=86.9.
deepseek-r1-0528,bfcl_v3_multiturn,37.0,https://huggingface.co/deepseek-ai/DeepSeek-R1-0528,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,DeepSeek R1-0528 model card: deepseek-r1-0528=37.0.
deepseek-r1-0528,tau_bench_airline,53.5,https://huggingface.co/deepseek-ai/DeepSeek-R1-0528,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,DeepSeek R1-0528 model card: deepseek-r1-0528=53.5.
deepseek-r1-0528,tau_bench_retail,63.9,https://huggingface.co/deepseek-ai/DeepSeek-R1-0528,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,DeepSeek R1-0528 model card: deepseek-r1-0528=63.9.
gemini-2.5-flash,aider_polyglot_whole,61.9,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-5-Flash-Model-Card.pdf,"{""effort"":""default"",""harness"":""AI Studio API"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""whole"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.5 Flash Model Card (Dec 2025) GA Thinking column: Aider Polyglot whole 61.9 (whole/diff: 61.9/56.7)
gemini-2.5-flash,facts_grounding,85.3,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-5-Flash-Model-Card.pdf,"{""effort"":""default"",""harness"":""AI Studio API"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.5 Flash Model Card (Dec 2025) GA Thinking column: FACTS Grounding 85.3
gemini-2.5-flash,vibe_eval,65.4,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-5-Flash-Model-Card.pdf,"{""effort"":""default"",""harness"":""AI Studio API"",""judge"":""Gemini"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.5 Flash Model Card (Dec 2025) GA Thinking column: Vibe-Eval (Reka) 65.4 (Gemini-as-judge)
gemini-2.5-flash,mrcr_v1,74.0,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-5-Flash-Model-Card.pdf,"{""effort"":""default"",""harness"":""AI Studio API"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.5 Flash Model Card (Dec 2025) GA Thinking column: MRCR (v1) 128k average 74
gemini-2.5-flash,global_mmlu_lite,88.4,https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-5-Flash-Model-Card.pdf,"{""effort"":""default"",""harness"":""AI Studio API"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Gemini 2.5 Flash Model Card (Dec 2025) GA Thinking column: Global MMLU Lite 88.4
deepseek-v2-0506,mmlu,78.2,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v2-0506=78.2.
deepseek-v2.5-0905,mmlu,80.6,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v2.5-0905=80.6.
qwen2.5-72b-instruct,mmlu,85.3,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: qwen2.5-72b-instruct=85.3.
llama-3.1-405b-instruct,mmlu,88.6,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: llama-3.1-405b-instruct=88.6.
claude-3.5-sonnet,mmlu,88.3,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: claude-3.5-sonnet=88.3.
gpt-4o-0513,mmlu,87.2,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: gpt-4o-0513=87.2.
deepseek-v2-0506,mmlu_redux,77.9,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v2-0506=77.9.
deepseek-v2.5-0905,mmlu_redux,80.3,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v2.5-0905=80.3.
qwen2.5-72b-instruct,mmlu_redux,85.6,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: qwen2.5-72b-instruct=85.6.
llama-3.1-405b-instruct,mmlu_redux,86.2,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: llama-3.1-405b-instruct=86.2.
claude-3.5-sonnet,mmlu_redux,88.9,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: claude-3.5-sonnet=88.9.
gpt-4o-0513,mmlu_redux,88.0,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: gpt-4o-0513=88.0.
deepseek-v3,mmlu_redux,89.1,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v3=89.1.
deepseek-v2-0506,mmlu_pro,58.5,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v2-0506=58.5.
deepseek-v2.5-0905,mmlu_pro,66.2,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v2.5-0905=66.2.
qwen2.5-72b-instruct,mmlu_pro,71.6,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: qwen2.5-72b-instruct=71.6.
llama-3.1-405b-instruct,mmlu_pro,73.3,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: llama-3.1-405b-instruct=73.3.
claude-3.5-sonnet,mmlu_pro,78.0,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: claude-3.5-sonnet=78.0.
gpt-4o-0513,mmlu_pro,72.6,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: gpt-4o-0513=72.6.
deepseek-v3,mmlu_pro,75.9,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v3=75.9.
deepseek-v2-0506,drop,83.0,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v2-0506=83.0.
deepseek-v2.5-0905,drop,87.8,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v2.5-0905=87.8.
qwen2.5-72b-instruct,drop,76.7,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: qwen2.5-72b-instruct=76.7.
llama-3.1-405b-instruct,drop,88.7,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: llama-3.1-405b-instruct=88.7.
claude-3.5-sonnet,drop,88.3,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: claude-3.5-sonnet=88.3.
gpt-4o-0513,drop,83.7,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: gpt-4o-0513=83.7.
deepseek-v3,drop,91.6,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v3=91.6.
deepseek-v2-0506,ifeval,57.7,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v2-0506=57.7.
deepseek-v2.5-0905,ifeval,80.6,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v2.5-0905=80.6.
qwen2.5-72b-instruct,ifeval,84.1,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: qwen2.5-72b-instruct=84.1.
llama-3.1-405b-instruct,ifeval,86.0,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: llama-3.1-405b-instruct=86.0.
claude-3.5-sonnet,ifeval,86.5,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: claude-3.5-sonnet=86.5.
gpt-4o-0513,ifeval,84.3,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: gpt-4o-0513=84.3.
deepseek-v2-0506,gpqa_diamond,35.3,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v2-0506=35.3.
deepseek-v2.5-0905,gpqa_diamond,41.3,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v2.5-0905=41.3.
qwen2.5-72b-instruct,gpqa_diamond,49.0,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: qwen2.5-72b-instruct=49.0.
llama-3.1-405b-instruct,gpqa_diamond,51.1,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: llama-3.1-405b-instruct=51.1.
claude-3.5-sonnet,gpqa_diamond,65.0,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: claude-3.5-sonnet=65.0.
gpt-4o-0513,gpqa_diamond,49.9,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: gpt-4o-0513=49.9.
deepseek-v2-0506,simpleqa,9.0,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v2-0506=9.0.
deepseek-v2.5-0905,simpleqa,10.2,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v2.5-0905=10.2.
qwen2.5-72b-instruct,simpleqa,9.1,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: qwen2.5-72b-instruct=9.1.
llama-3.1-405b-instruct,simpleqa,17.1,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: llama-3.1-405b-instruct=17.1.
claude-3.5-sonnet,simpleqa,28.4,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: claude-3.5-sonnet=28.4.
gpt-4o-0513,simpleqa,38.2,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: gpt-4o-0513=38.2.
deepseek-v3,simpleqa,24.9,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v3=24.9.
deepseek-v2-0506,frames,66.9,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v2-0506=66.9.
deepseek-v2.5-0905,frames,65.4,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v2.5-0905=65.4.
qwen2.5-72b-instruct,frames,69.8,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: qwen2.5-72b-instruct=69.8.
llama-3.1-405b-instruct,frames,70.0,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: llama-3.1-405b-instruct=70.0.
claude-3.5-sonnet,frames,72.5,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: claude-3.5-sonnet=72.5.
gpt-4o-0513,frames,80.5,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: gpt-4o-0513=80.5.
deepseek-v3,frames,73.3,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v3=73.3.
deepseek-v2-0506,longbench_v2,31.6,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v2-0506=31.6.
deepseek-v2.5-0905,longbench_v2,35.4,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v2.5-0905=35.4.
qwen2.5-72b-instruct,longbench_v2,39.4,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: qwen2.5-72b-instruct=39.4.
llama-3.1-405b-instruct,longbench_v2,36.1,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: llama-3.1-405b-instruct=36.1.
claude-3.5-sonnet,longbench_v2,41.0,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: claude-3.5-sonnet=41.0.
gpt-4o-0513,longbench_v2,48.1,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: gpt-4o-0513=48.1.
deepseek-v3,longbench_v2,48.7,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v3=48.7.
deepseek-v2-0506,humaneval,69.3,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v2-0506=69.3.
deepseek-v2.5-0905,humaneval,77.4,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v2.5-0905=77.4.
qwen2.5-72b-instruct,humaneval,77.3,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: qwen2.5-72b-instruct=77.3.
llama-3.1-405b-instruct,humaneval,77.2,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: llama-3.1-405b-instruct=77.2.
claude-3.5-sonnet,humaneval,81.7,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: claude-3.5-sonnet=81.7.
gpt-4o-0513,humaneval,80.5,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: gpt-4o-0513=80.5.
deepseek-v2-0506,livecodebench,18.8,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v2-0506=18.8.
deepseek-v2.5-0905,livecodebench,29.2,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v2.5-0905=29.2.
qwen2.5-72b-instruct,livecodebench,31.1,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: qwen2.5-72b-instruct=31.1.
llama-3.1-405b-instruct,livecodebench,28.4,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: llama-3.1-405b-instruct=28.4.
claude-3.5-sonnet,livecodebench,36.3,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: claude-3.5-sonnet=36.3.
gpt-4o-0513,livecodebench,33.4,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: gpt-4o-0513=33.4.
deepseek-v3,livecodebench,40.5,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v3=40.5.
deepseek-v2.5-0905,swe_bench_verified,22.6,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v2.5-0905=22.6.
qwen2.5-72b-instruct,swe_bench_verified,23.8,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,tech_report,verified,DeepSeek V3 paper Table 6: qwen2.5-72b-instruct=23.8.
llama-3.1-405b-instruct,swe_bench_verified,24.5,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,tech_report,verified,DeepSeek V3 paper Table 6: llama-3.1-405b-instruct=24.5.
claude-3.5-sonnet,swe_bench_verified,50.8,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,tech_report,verified,DeepSeek V3 paper Table 6: claude-3.5-sonnet=50.8.
gpt-4o-0513,swe_bench_verified,38.8,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,tech_report,verified,DeepSeek V3 paper Table 6: gpt-4o-0513=38.8.
deepseek-v2.5-0905,aider_polyglot_diff,18.2,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v2.5-0905=18.2.
qwen2.5-72b-instruct,aider_polyglot_diff,7.6,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,tech_report,verified,DeepSeek V3 paper Table 6: qwen2.5-72b-instruct=7.6.
llama-3.1-405b-instruct,aider_polyglot_diff,5.8,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,tech_report,verified,DeepSeek V3 paper Table 6: llama-3.1-405b-instruct=5.8.
claude-3.5-sonnet,aider_polyglot_diff,45.3,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,tech_report,verified,DeepSeek V3 paper Table 6: claude-3.5-sonnet=45.3.
gpt-4o-0513,aider_polyglot_diff,16.0,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,tech_report,verified,DeepSeek V3 paper Table 6: gpt-4o-0513=16.0.
deepseek-v3,aider_polyglot_diff,49.6,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v3=49.6.
deepseek-v2-0506,aime_2024,4.6,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v2-0506=4.6.
deepseek-v2.5-0905,aime_2024,16.7,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v2.5-0905=16.7.
qwen2.5-72b-instruct,aime_2024,23.3,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: qwen2.5-72b-instruct=23.3.
llama-3.1-405b-instruct,aime_2024,23.3,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: llama-3.1-405b-instruct=23.3.
claude-3.5-sonnet,aime_2024,16.0,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: claude-3.5-sonnet=16.0.
gpt-4o-0513,aime_2024,9.3,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: gpt-4o-0513=9.3.
deepseek-v3,aime_2024,39.2,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v3=39.2.
deepseek-v2-0506,math_500,56.3,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v2-0506=56.3.
deepseek-v2.5-0905,math_500,74.7,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v2.5-0905=74.7.
qwen2.5-72b-instruct,math_500,80.0,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: qwen2.5-72b-instruct=80.0.
llama-3.1-405b-instruct,math_500,73.8,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: llama-3.1-405b-instruct=73.8.
claude-3.5-sonnet,math_500,78.3,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: claude-3.5-sonnet=78.3.
gpt-4o-0513,math_500,74.6,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: gpt-4o-0513=74.6.
deepseek-v2-0506,cnmo_2024,2.8,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v2-0506=2.8.
deepseek-v2.5-0905,cnmo_2024,10.8,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v2.5-0905=10.8.
qwen2.5-72b-instruct,cnmo_2024,15.9,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: qwen2.5-72b-instruct=15.9.
llama-3.1-405b-instruct,cnmo_2024,6.8,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: llama-3.1-405b-instruct=6.8.
claude-3.5-sonnet,cnmo_2024,13.1,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: claude-3.5-sonnet=13.1.
gpt-4o-0513,cnmo_2024,10.8,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: gpt-4o-0513=10.8.
deepseek-v3,cnmo_2024,43.2,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v3=43.2.
deepseek-v2-0506,cluewsc,89.9,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v2-0506=89.9.
deepseek-v2.5-0905,cluewsc,90.4,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v2.5-0905=90.4.
qwen2.5-72b-instruct,cluewsc,91.4,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: qwen2.5-72b-instruct=91.4.
llama-3.1-405b-instruct,cluewsc,84.7,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: llama-3.1-405b-instruct=84.7.
claude-3.5-sonnet,cluewsc,85.4,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: claude-3.5-sonnet=85.4.
gpt-4o-0513,cluewsc,87.9,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: gpt-4o-0513=87.9.
deepseek-v3,cluewsc,90.9,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v3=90.9.
deepseek-v2-0506,c_eval,78.6,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v2-0506=78.6.
deepseek-v2.5-0905,c_eval,79.5,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v2.5-0905=79.5.
qwen2.5-72b-instruct,c_eval,86.1,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: qwen2.5-72b-instruct=86.1.
llama-3.1-405b-instruct,c_eval,61.5,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: llama-3.1-405b-instruct=61.5.
claude-3.5-sonnet,c_eval,76.7,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: claude-3.5-sonnet=76.7.
gpt-4o-0513,c_eval,76.0,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: gpt-4o-0513=76.0.
deepseek-v3,c_eval,86.5,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v3=86.5.
deepseek-v2-0506,chinese_simpleqa,48.5,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v2-0506=48.5.
deepseek-v2.5-0905,chinese_simpleqa,54.1,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v2.5-0905=54.1.
qwen2.5-72b-instruct,chinese_simpleqa,48.4,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: qwen2.5-72b-instruct=48.4.
llama-3.1-405b-instruct,chinese_simpleqa,50.4,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: llama-3.1-405b-instruct=50.4.
claude-3.5-sonnet,chinese_simpleqa,51.3,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: claude-3.5-sonnet=51.3.
gpt-4o-0513,chinese_simpleqa,59.3,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: gpt-4o-0513=59.3.
deepseek-v3,chinese_simpleqa,64.8,https://arxiv.org/abs/2412.19437,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek V3 paper Table 6: deepseek-v3=64.8.
deepseek-r1,drop,92.2,https://arxiv.org/abs/2501.12948,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,"DeepSeek R1 paper Table 3 (page 9, R1 final): 92.2."
deepseek-r1,alpacaeval_2,87.6,https://arxiv.org/abs/2501.12948,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,"DeepSeek R1 paper Table 3 (page 9, R1 final): 87.6."
deepseek-r1,cluewsc,92.8,https://arxiv.org/abs/2501.12948,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,"DeepSeek R1 paper Table 3 (page 9, R1 final): 92.8."
deepseek-r1,c_eval,91.8,https://arxiv.org/abs/2501.12948,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,"DeepSeek R1 paper Table 3 (page 9, R1 final): 91.8."
deepseek-r1,chinese_simpleqa,63.7,https://arxiv.org/abs/2501.12948,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,"DeepSeek R1 paper Table 3 (page 9, R1 final): 63.7."
gpt-4o-0513,codeforces_rating,759.0,https://arxiv.org/abs/2501.12948,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek R1 paper Table 15: gpt-4o-0513=759.
claude-3.5-sonnet,codeforces_rating,717.0,https://arxiv.org/abs/2501.12948,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,DeepSeek R1 paper Table 15: claude-3.5-sonnet=717.
qwq-32b,livebench,72.0,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,QwQ-32B blog (qwenlm.github.io/blog/qwq-32b/) + Qwen3 tech report Table 13 baseline column: LiveBench 2024-11-25 72.0
qwq-32b,bfcl_v3,66.4,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,QwQ-32B blog (qwenlm.github.io/blog/qwq-32b/) + Qwen3 tech report Table 13 baseline column: BFCL v3 66.4
qwq-32b,multi_if,68.3,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,QwQ-32B blog (qwenlm.github.io/blog/qwq-32b/) + Qwen3 tech report Table 13 baseline column: Multi-IF 68.3
qwq-32b,mmmlu,80.9,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,QwQ-32B blog (qwenlm.github.io/blog/qwq-32b/) + Qwen3 tech report Table 13 baseline column: MMMLU 14 langs 80.9
qwq-32b,mt_aime_2024,68.0,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,QwQ-32B blog (qwenlm.github.io/blog/qwq-32b/) + Qwen3 tech report Table 13 baseline column: MT-AIME2024 68.0
qwq-32b,mmlu_redux,90.0,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,QwQ-32B blog (qwenlm.github.io/blog/qwq-32b/) + Qwen3 tech report Table 13 baseline column: MMLU-Redux 90.0
deepseek-r1-distill-llama-70b,ifeval,79.3,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Qwen3 tech report (arxiv:2505.09388) Table 13 baseline column for DeepSeek-R1-Distill-Llama-70B (third-party Qwen self-test, matches_canonical=false): IFEval strict prompt 79.3"
deepseek-r1-distill-llama-70b,livebench,54.5,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Qwen3 tech report (arxiv:2505.09388) Table 13 baseline column for DeepSeek-R1-Distill-Llama-70B (third-party Qwen self-test, matches_canonical=false): LiveBench 2024-11-25 54.5"
deepseek-r1-distill-llama-70b,arena_hard,60.6,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Qwen3 tech report (arxiv:2505.09388) Table 13 baseline column for DeepSeek-R1-Distill-Llama-70B (third-party Qwen self-test, matches_canonical=false): Arena-Hard 60.6"
deepseek-r1-distill-llama-70b,mmlu_redux,89.3,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Qwen3 tech report (arxiv:2505.09388) Table 13 baseline column for DeepSeek-R1-Distill-Llama-70B (third-party Qwen self-test, matches_canonical=false): MMLU-Redux 89.3"
deepseek-r1-distill-llama-70b,bfcl_v3,49.3,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Qwen3 tech report (arxiv:2505.09388) Table 13 baseline column for DeepSeek-R1-Distill-Llama-70B (third-party Qwen self-test, matches_canonical=false): BFCL v3 49.3"
deepseek-r1-distill-llama-70b,multi_if,57.6,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Qwen3 tech report (arxiv:2505.09388) Table 13 baseline column for DeepSeek-R1-Distill-Llama-70B (third-party Qwen self-test, matches_canonical=false): Multi-IF 57.6"
deepseek-r1-distill-llama-70b,mmmlu,69.6,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Qwen3 tech report (arxiv:2505.09388) Table 13 baseline column for DeepSeek-R1-Distill-Llama-70B (third-party Qwen self-test, matches_canonical=false): MMMLU 14 langs 69.6"
deepseek-r1-distill-llama-70b,mt_aime_2024,29.3,https://arxiv.org/abs/2505.09388,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Qwen3 tech report (arxiv:2505.09388) Table 13 baseline column for DeepSeek-R1-Distill-Llama-70B (third-party Qwen self-test, matches_canonical=false): MT-AIME2024 29.3"
glm-5,hle,30.5,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5.1 blog: glm-5=30.5.
minimax-m2.7,hle,28.0,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5.1 blog: minimax-m2.7=28.0.
minimax-m2.7,aime_2026,89.8,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5.1 blog: minimax-m2.7=89.8.
claude-opus-4.6,aime_2026,95.6,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5.1 blog: claude-opus-4.6=95.6.
gpt-5.4,aime_2026,98.7,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5.1 blog: gpt-5.4=98.7.
glm-5.1,hmmt_nov_2025,94.0,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5.1 blog: glm-5.1=94.0.
glm-5,hmmt_nov_2025,96.9,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5.1 blog: glm-5=96.9.
qwen3.6-plus,hmmt_nov_2025,94.6,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5.1 blog: qwen3.6-plus=94.6.
minimax-m2.7,hmmt_nov_2025,81.0,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5.1 blog: minimax-m2.7=81.0.
claude-opus-4.6,hmmt_nov_2025,96.3,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5.1 blog: claude-opus-4.6=96.3.
gemini-3.1-pro,hmmt_nov_2025,94.8,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5.1 blog: gemini-3.1-pro=94.8.
gpt-5.4,hmmt_nov_2025,95.8,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5.1 blog: gpt-5.4=95.8.
qwen3.6-plus,hmmt_feb_2026,87.8,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5.1 blog: qwen3.6-plus=87.8.
minimax-m2.7,hmmt_feb_2026,72.7,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5.1 blog: minimax-m2.7=72.7.
glm-5,imo_answerbench,82.5,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5.1 blog: glm-5=82.5.
qwen3.6-plus,imo_answerbench,83.8,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5.1 blog: qwen3.6-plus=83.8.
minimax-m2.7,imo_answerbench,66.3,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5.1 blog: minimax-m2.7=66.3.
glm-5,gpqa_diamond,86.0,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5.1 blog: glm-5=86.0.
minimax-m2.7,gpqa_diamond,87.0,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5.1 blog: minimax-m2.7=87.0.
glm-5,swe_bench_pro,55.1,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,GLM-5.1 blog: glm-5=55.1.
minimax-m2.7,swe_bench_pro,56.2,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,GLM-5.1 blog: minimax-m2.7=56.2.
glm-5.1,nl2repo_bench,42.7,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,GLM-5.1 blog: glm-5.1=42.7.
glm-5,nl2repo_bench,35.9,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,GLM-5.1 blog: glm-5=35.9.
qwen3.6-plus,nl2repo_bench,37.9,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,GLM-5.1 blog: qwen3.6-plus=37.9.
minimax-m2.7,nl2repo_bench,39.8,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,GLM-5.1 blog: minimax-m2.7=39.8.
kimi-k2.5,nl2repo_bench,32.0,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,GLM-5.1 blog: kimi-k2.5=32.0.
claude-opus-4.6,nl2repo_bench,49.8,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,GLM-5.1 blog: claude-opus-4.6=49.8.
gemini-3.1-pro,nl2repo_bench,33.4,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,GLM-5.1 blog: gemini-3.1-pro=33.4.
gpt-5.4,nl2repo_bench,41.3,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,GLM-5.1 blog: gpt-5.4=41.3.
glm-5,terminal_bench,56.2,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5.1 blog: glm-5=56.2.
glm-5.1,cybergym,68.7,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5.1 blog: glm-5.1=68.7.
glm-5,cybergym,48.3,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5.1 blog: glm-5=48.3.
deepseek-v3.2,cybergym,17.3,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5.1 blog: deepseek-v3.2=17.3.
gemini-3.1-pro,cybergym,38.8,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5.1 blog: gemini-3.1-pro=38.8.
glm-5,browsecomp,62.0,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5.1 blog: glm-5=62.0.
glm-5.1,tau3_bench,70.6,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5.1 blog: glm-5.1=70.6.
glm-5,tau3_bench,69.2,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5.1 blog: glm-5=69.2.
qwen3.6-plus,tau3_bench,70.7,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5.1 blog: qwen3.6-plus=70.7.
minimax-m2.7,tau3_bench,67.6,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5.1 blog: minimax-m2.7=67.6.
deepseek-v3.2,tau3_bench,69.2,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5.1 blog: deepseek-v3.2=69.2.
kimi-k2.5,tau3_bench,66.0,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5.1 blog: kimi-k2.5=66.0.
claude-opus-4.6,tau3_bench,72.4,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5.1 blog: claude-opus-4.6=72.4.
gemini-3.1-pro,tau3_bench,67.1,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5.1 blog: gemini-3.1-pro=67.1.
gpt-5.4,tau3_bench,72.9,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5.1 blog: gpt-5.4=72.9.
glm-5,mcpatlas,69.2,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5.1 blog: glm-5=69.2.
qwen3.6-plus,mcpatlas,74.1,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5.1 blog: qwen3.6-plus=74.1.
minimax-m2.7,mcpatlas,48.8,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5.1 blog: minimax-m2.7=48.8.
deepseek-v3.2,mcpatlas,62.2,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5.1 blog: deepseek-v3.2=62.2.
kimi-k2.5,mcpatlas,63.8,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5.1 blog: kimi-k2.5=63.8.
glm-5,toolathlon,38.0,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5.1 blog: glm-5=38.0.
qwen3.6-plus,toolathlon,39.8,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5.1 blog: qwen3.6-plus=39.8.
minimax-m2.7,toolathlon,46.3,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5.1 blog: minimax-m2.7=46.3.
deepseek-v3.2,toolathlon,35.2,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5.1 blog: deepseek-v3.2=35.2.
glm-5.1,vending_bench_2,5634.41,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5.1 blog: glm-5.1=5634.41.
glm-5,vending_bench_2,4432.12,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5.1 blog: glm-5=4432.12.
qwen3.6-plus,vending_bench_2,5114.87,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5.1 blog: qwen3.6-plus=5114.87.
deepseek-v3.2,vending_bench_2,1034.0,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5.1 blog: deepseek-v3.2=1034.0.
kimi-k2.5,vending_bench_2,1198.46,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5.1 blog: kimi-k2.5=1198.46.
claude-opus-4.6,vending_bench_2,8017.59,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5.1 blog: claude-opus-4.6=8017.59.
gemini-3.1-pro,vending_bench_2,911.21,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5.1 blog: gemini-3.1-pro=911.21.
gpt-5.4,vending_bench_2,6144.18,https://z.ai/blog/glm-5.1,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5.1 blog: gpt-5.4=6144.18.
glm-4.7,hmmt_nov_2025,93.5,https://z.ai/blog/glm-5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5 blog: glm-4.7=93.5.
claude-opus-4.5,hmmt_nov_2025,91.7,https://z.ai/blog/glm-5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5 blog: claude-opus-4.5=91.7.
glm-4.7,imo_answerbench,82.0,https://z.ai/blog/glm-5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5 blog: glm-4.7=82.0.
claude-opus-4.5,imo_answerbench,78.5,https://z.ai/blog/glm-5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5 blog: claude-opus-4.5=78.5.
gpt-5.2,imo_answerbench,86.3,https://z.ai/blog/glm-5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-5 blog: gpt-5.2=86.3.
glm-5,swe_bench_verified,77.8,https://z.ai/blog/glm-5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,GLM-5 blog: glm-5=77.8.
glm-5,swe_bench_multilingual,73.3,https://z.ai/blog/glm-5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,GLM-5 blog: glm-5=73.3.
glm-4.7,swe_bench_multilingual,66.7,https://z.ai/blog/glm-5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,GLM-5 blog: glm-4.7=66.7.
claude-opus-4.5,swe_bench_multilingual,77.5,https://z.ai/blog/glm-5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,GLM-5 blog: claude-opus-4.5=77.5.
gemini-3-pro,swe_bench_multilingual,65.0,https://z.ai/blog/glm-5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,GLM-5 blog: gemini-3-pro=65.0.
gpt-5.2,swe_bench_multilingual,72.0,https://z.ai/blog/glm-5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,GLM-5 blog: gpt-5.2=72.0.
glm-4.7,cybergym,23.5,https://z.ai/blog/glm-5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5 blog: glm-4.7=23.5.
claude-opus-4.5,cybergym,50.6,https://z.ai/blog/glm-5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5 blog: claude-opus-4.5=50.6.
gemini-3-pro,cybergym,39.9,https://z.ai/blog/glm-5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5 blog: gemini-3-pro=39.9.
glm-5,browsecomp_zh,72.7,https://z.ai/blog/glm-5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5 blog: glm-5=72.7.
glm-4.7,browsecomp_zh,66.6,https://z.ai/blog/glm-5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5 blog: glm-4.7=66.6.
deepseek-v3.2,browsecomp_zh,65.0,https://z.ai/blog/glm-5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5 blog: deepseek-v3.2=65.0.
kimi-k2.5,browsecomp_zh,62.3,https://z.ai/blog/glm-5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5 blog: kimi-k2.5=62.3.
claude-opus-4.5,browsecomp_zh,62.4,https://z.ai/blog/glm-5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5 blog: claude-opus-4.5=62.4.
gemini-3-pro,browsecomp_zh,66.8,https://z.ai/blog/glm-5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5 blog: gemini-3-pro=66.8.
gpt-5.2,browsecomp_zh,76.1,https://z.ai/blog/glm-5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5 blog: gpt-5.2=76.1.
glm-4.7,mcpatlas,52.0,https://z.ai/blog/glm-5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5 blog: glm-4.7=52.0.
glm-4.7,toolathlon,23.8,https://z.ai/blog/glm-5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5 blog: glm-4.7=23.8.
claude-opus-4.5,toolathlon,43.5,https://z.ai/blog/glm-5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5 blog: claude-opus-4.5=43.5.
glm-4.7,vending_bench_2,2376.82,https://z.ai/blog/glm-5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5 blog: glm-4.7=2376.82.
claude-opus-4.5,vending_bench_2,4967.06,https://z.ai/blog/glm-5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5 blog: claude-opus-4.5=4967.06.
glm-4.7,hmmt_feb_2025,97.1,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-4.7 blog: glm-4.7=97.1.
glm-4.6,hmmt_feb_2025,89.2,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-4.7 blog: glm-4.6=89.2.
glm-4.6,hmmt_nov_2025,87.7,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-4.7 blog: glm-4.6=87.7.
glm-4.6,imo_answerbench,73.5,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-4.7 blog: glm-4.6=73.5.
claude-sonnet-4.5,imo_answerbench,65.8,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-4.7 blog: claude-sonnet-4.5=65.8.
glm-4.7,livecodebench_v6,84.9,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-4.7 blog: glm-4.7=84.9.
glm-4.6,livecodebench_v6,82.8,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-4.7 blog: glm-4.6=82.8.
kimi-k2-thinking,livecodebench_v6,83.1,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-4.7 blog: kimi-k2-thinking=83.1.
deepseek-v3.2,livecodebench_v6,83.3,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-4.7 blog: deepseek-v3.2=83.3.
gemini-3-pro,livecodebench_v6,90.7,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-4.7 blog: gemini-3-pro=90.7.
claude-sonnet-4.5,livecodebench_v6,64.0,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-4.7 blog: claude-sonnet-4.5=64.0.
gpt-5,livecodebench_v6,87.0,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-4.7 blog: gpt-5=87.0.
gpt-5.1,livecodebench_v6,87.0,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,GLM-4.7 blog: gpt-5.1=87.0.
glm-4.6,swe_bench_multilingual,53.8,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,official_blog,verified,GLM-4.7 blog: glm-4.6=53.8.
glm-4.7,terminal_bench_hard,33.3,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-4.7 blog: glm-4.7=33.3.
glm-4.6,terminal_bench_hard,23.6,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-4.7 blog: glm-4.6=23.6.
kimi-k2-thinking,terminal_bench_hard,30.6,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-4.7 blog: kimi-k2-thinking=30.6.
deepseek-v3.2,terminal_bench_hard,35.4,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-4.7 blog: deepseek-v3.2=35.4.
gemini-3-pro,terminal_bench_hard,39.0,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-4.7 blog: gemini-3-pro=39.0.
claude-sonnet-4.5,terminal_bench_hard,33.3,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-4.7 blog: claude-sonnet-4.5=33.3.
gpt-5,terminal_bench_hard,30.5,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-4.7 blog: gpt-5=30.5.
gpt-5.1,terminal_bench_hard,43.0,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-4.7 blog: gpt-5.1=43.0.
glm-4.6,browsecomp,45.1,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-4.7 blog: glm-4.6=45.1.
glm-4.6,browsecomp_zh,49.5,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-4.7 blog: glm-4.6=49.5.
claude-sonnet-4.5,browsecomp_zh,42.4,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-4.7 blog: claude-sonnet-4.5=42.4.
gpt-5,browsecomp_zh,63.0,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-4.7 blog: gpt-5=63.0.
kimi-k2.6,deepsearchqa_f1,92.5,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.6 model card: kimi-k2.6=92.5.
gpt-5.4,deepsearchqa_f1,78.6,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.6 model card: gpt-5.4=78.6.
claude-opus-4.6,deepsearchqa_f1,91.3,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.6 model card: claude-opus-4.6=91.3.
gemini-3.1-pro,deepsearchqa_f1,81.9,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.6 model card: gemini-3.1-pro=81.9.
kimi-k2.5,deepsearchqa_f1,89.0,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.6 model card: kimi-k2.5=89.0.
kimi-k2.6,deepsearchqa_acc,83.0,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.6 model card: kimi-k2.6=83.0.
gpt-5.4,deepsearchqa_acc,63.7,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.6 model card: gpt-5.4=63.7.
claude-opus-4.6,deepsearchqa_acc,80.6,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.6 model card: claude-opus-4.6=80.6.
gemini-3.1-pro,deepsearchqa_acc,60.2,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.6 model card: gemini-3.1-pro=60.2.
kimi-k2.5,deepsearchqa_acc,77.1,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.6 model card: kimi-k2.5=77.1.
kimi-k2.6,widesearch,80.8,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.6 model card: kimi-k2.6=80.8.
kimi-k2.5,widesearch,72.7,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.6 model card: kimi-k2.5=72.7.
gpt-5.4,mcpmark,62.5,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.6 model card: gpt-5.4=62.5.
claude-opus-4.6,mcpmark,56.7,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.6 model card: claude-opus-4.6=56.7.
gemini-3.1-pro,mcpmark,55.9,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.6 model card: gemini-3.1-pro=55.9.
kimi-k2.5,mcpmark,29.5,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.6 model card: kimi-k2.5=29.5.
kimi-k2.6,apex_agents,27.9,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.6 model card: kimi-k2.6=27.9.
gpt-5.4,apex_agents,33.3,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.6 model card: gpt-5.4=33.3.
kimi-k2.5,apex_agents,11.5,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.6 model card: kimi-k2.5=11.5.
kimi-k2.6,osworld,73.1,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.6 model card: kimi-k2.6=73.1.
gemini-3.1-pro,swe_bench_multilingual,76.9,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic (bash + edit)""}",True,model_card,verified,Kimi K2.6 model card: gemini-3.1-pro=76.9.
gpt-5.4,scicode,56.6,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.6 model card: gpt-5.4=56.6.
kimi-k2.6,ojbench,60.6,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.6 model card: kimi-k2.6=60.6.
claude-opus-4.6,ojbench,60.3,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.6 model card: claude-opus-4.6=60.3.
gemini-3.1-pro,ojbench,70.7,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.6 model card: gemini-3.1-pro=70.7.
kimi-k2.6,livecodebench_v6,89.6,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.6 model card: kimi-k2.6=89.6.
claude-opus-4.6,livecodebench_v6,88.8,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.6 model card: claude-opus-4.6=88.8.
gemini-3.1-pro,livecodebench_v6,91.7,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.6 model card: gemini-3.1-pro=91.7.
kimi-k2.5,livecodebench_v6,85.0,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.6 model card: kimi-k2.5=85.0.
gpt-5.4,charxiv_reasoning,82.8,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.6 model card: gpt-5.4=82.8.
gemini-3.1-pro,charxiv_reasoning,80.2,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.6 model card: gemini-3.1-pro=80.2.
gpt-5.4,mathvision,92.0,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.6 model card: gpt-5.4=92.0.
claude-opus-4.6,mathvision,71.2,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.6 model card: claude-opus-4.6=71.2.
gpt-5.4,babyvision,49.7,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.6 model card: gpt-5.4=49.7.
claude-opus-4.6,babyvision,14.8,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.6 model card: claude-opus-4.6=14.8.
gemini-3.1-pro,babyvision,51.6,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.6 model card: gemini-3.1-pro=51.6.
kimi-k2.5,babyvision,36.5,https://huggingface.co/moonshotai/Kimi-K2.6,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.6 model card: kimi-k2.5=36.5.
qwen3.5-397b,supergpqa,70.4,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Qwen3.5-397B-A17B HF model card cross-model table: SuperGPQA 70.4
qwen3.5-397b,mmlu_redux,94.9,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Qwen3.5-397B-A17B HF model card cross-model table: MMLU-Redux 94.9
qwen3.5-397b,multichallenge,67.6,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Qwen3.5-397B-A17B HF model card cross-model table: MultiChallenge 67.6
qwen3.5-397b,longbench_v2,63.2,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Qwen3.5-397B-A17B HF model card cross-model table: LongBench v2 63.2
qwen3.5-397b,mcpmark,46.1,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Qwen3.5-397B-A17B HF model card cross-model table: MCP-Mark 46.1
qwen3.5-397b,toolathlon,38.3,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Qwen3.5-397B-A17B HF model card cross-model table: Tool Decathlon 38.3
qwen3.5-397b,browsecomp_zh,70.3,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Qwen3.5-397B-A17B HF model card cross-model table: BrowseComp-zh 70.3
qwen3.5-397b,global_piqa,89.8,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Qwen3.5-397B-A17B HF model card cross-model table: Global PIQA 89.8
qwen3.5-397b,mmmu_pro,79.0,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Qwen3.5-397B-A17B HF model card cross-model table: MMMU-Pro 79.0
qwen3.5-397b,mathvista_mini,90.3,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Qwen3.5-397B-A17B HF model card cross-model table: MathVista (mini) 90.3
qwen3.5-397b,charxiv_reasoning,80.8,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Qwen3.5-397B-A17B HF model card cross-model table: CharXiv RQ 80.8
qwen3.5-397b,ocrbench,93.1,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Qwen3.5-397B-A17B HF model card cross-model table: OCRBench 93.1
qwen3.5-397b,screenspot_pro,65.6,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Qwen3.5-397B-A17B HF model card cross-model table: ScreenSpot Pro 65.6
qwen3.5-397b,video_mme,87.5,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Qwen3.5-397B-A17B HF model card cross-model table: Video-MME with subtitle 87.5
qwen3.5-397b,video_mmmu,84.7,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Qwen3.5-397B-A17B HF model card cross-model table: Video-MMU 84.7
kimi-k2.5,hmmt_feb_2025,95.4,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: kimi-k2.5=95.4.
claude-opus-4.5,hmmt_feb_2025,92.9,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: claude-opus-4.5=92.9.
claude-opus-4.5,charxiv_reasoning,67.2,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: claude-opus-4.5=67.2.
gpt-5.2,mathvision,83.0,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: gpt-5.2=83.0.
claude-opus-4.5,mathvision,77.1,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: claude-opus-4.5=77.1.
gemini-3-pro,mathvision,86.1,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: gemini-3-pro=86.1.
gpt-5.2,mathvista_mini,82.8,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: gpt-5.2=82.8.
claude-opus-4.5,mathvista_mini,80.2,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: claude-opus-4.5=80.2.
gemini-3-pro,mathvista_mini,89.8,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: gemini-3-pro=89.8.
kimi-k2.5,zerobench_main,9.0,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: kimi-k2.5=9.
gpt-5.2,ocrbench,80.7,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: gpt-5.2=80.7.
claude-opus-4.5,ocrbench,86.5,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: claude-opus-4.5=86.5.
gemini-3-pro,ocrbench,90.3,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: gemini-3-pro=90.3.
kimi-k2.5,omnidocbench_1.5,0.112,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: kimi-k2.5=88.8. [Converted 88.8% (1-LED higher-better) → 0.112 (edit distance lower-better) for cross-source consistency]
gpt-5.2,omnidocbench_1.5,0.143,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: gpt-5.2=85.7. [Converted 85.7% (1-LED higher-better) → 0.143 (edit distance lower-better) for cross-source consistency]
claude-opus-4.5,omnidocbench_1.5,0.123,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: claude-opus-4.5=87.7. [Converted 87.7% (1-LED higher-better) → 0.123 (edit distance lower-better) for cross-source consistency]
gemini-3-pro,omnidocbench_1.5,0.115,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: gemini-3-pro=88.5. [Converted 88.5% (1-LED higher-better) → 0.115 (edit distance lower-better) for cross-source consistency]
kimi-k2.5,infovqa,92.6,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: kimi-k2.5=92.6.
gpt-5.2,infovqa,84.0,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: gpt-5.2=84.
claude-opus-4.5,infovqa,76.9,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: claude-opus-4.5=76.9.
gemini-3-pro,infovqa,57.2,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: gemini-3-pro=57.2.
kimi-k2.5,simplevqa,71.2,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: kimi-k2.5=71.2.
gpt-5.2,simplevqa,55.8,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: gpt-5.2=55.8.
claude-opus-4.5,simplevqa,69.7,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: claude-opus-4.5=69.7.
gemini-3-pro,simplevqa,69.7,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: gemini-3-pro=69.7.
kimi-k2.5,worldvqa,46.3,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: kimi-k2.5=46.3.
gpt-5.2,worldvqa,28.0,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: gpt-5.2=28.0.
claude-opus-4.5,worldvqa,36.8,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: claude-opus-4.5=36.8.
gemini-3-pro,worldvqa,47.4,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: gemini-3-pro=47.4.
kimi-k2.5,mmvu,80.4,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: kimi-k2.5=80.4.
gpt-5.2,mmvu,80.8,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: gpt-5.2=80.8.
claude-opus-4.5,mmvu,77.3,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: claude-opus-4.5=77.3.
gemini-3-pro,mmvu,77.5,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: gemini-3-pro=77.5.
kimi-k2.5,motionbench,70.4,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: kimi-k2.5=70.4.
gpt-5.2,motionbench,64.8,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: gpt-5.2=64.8.
claude-opus-4.5,motionbench,60.3,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: claude-opus-4.5=60.3.
gemini-3-pro,motionbench,70.3,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: gemini-3-pro=70.3.
gpt-5.2,video_mme,86.0,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: gpt-5.2=86.0.
gemini-3-pro,video_mme,88.4,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: gemini-3-pro=88.4.
kimi-k2.5,longvideobench,79.8,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: kimi-k2.5=79.8.
gpt-5.2,longvideobench,76.5,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: gpt-5.2=76.5.
claude-opus-4.5,longvideobench,67.2,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: claude-opus-4.5=67.2.
gemini-3-pro,longvideobench,77.7,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: gemini-3-pro=77.7.
kimi-k2.5,lvbench,75.9,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: kimi-k2.5=75.9.
gemini-3-pro,lvbench,73.5,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: gemini-3-pro=73.5.
gpt-5.2,paperbench,63.7,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.5 model card: gpt-5.2=63.7.
claude-opus-4.5,paperbench,72.9,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.5 model card: claude-opus-4.5=72.9.
deepseek-v3.2,paperbench,47.1,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.5 model card: deepseek-v3.2=47.1.
claude-opus-4.5,scicode,49.5,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: claude-opus-4.5=49.5.
deepseek-v3.2,scicode,38.9,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: deepseek-v3.2=38.9.
claude-opus-4.5,ojbench,54.6,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: claude-opus-4.5=54.6.
gemini-3-pro,ojbench,68.5,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: gemini-3-pro=68.5.
deepseek-v3.2,ojbench,54.7,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: deepseek-v3.2=54.7.
claude-opus-4.5,livecodebench_v6,82.2,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: claude-opus-4.5=82.2.
gpt-5.2,longbench_v2,54.5,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: gpt-5.2=54.5.
claude-opus-4.5,longbench_v2,64.4,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: claude-opus-4.5=64.4.
gemini-3-pro,longbench_v2,68.2,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: gemini-3-pro=68.2.
deepseek-v3.2,longbench_v2,59.8,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: deepseek-v3.2=59.8.
gemini-3-pro,aa_lcr,65.3,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: gemini-3-pro=65.3.
deepseek-v3.2,aa_lcr,64.3,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2.5 model card: deepseek-v3.2=64.3.
claude-opus-4.5,widesearch,76.2,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.5 model card: claude-opus-4.5=76.2.
gemini-3-pro,widesearch,57.0,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.5 model card: gemini-3-pro=57.0.
gpt-5.2,deepsearchqa_acc,71.3,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.5 model card: gpt-5.2=71.3.
claude-opus-4.5,deepsearchqa_acc,76.1,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.5 model card: claude-opus-4.5=76.1.
gemini-3-pro,deepsearchqa_acc,63.2,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.5 model card: gemini-3-pro=63.2.
deepseek-v3.2,deepsearchqa_acc,60.9,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.5 model card: deepseek-v3.2=60.9.
kimi-k2.5,finsearchcompt23,67.8,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.5 model card: kimi-k2.5=67.8.
claude-opus-4.5,finsearchcompt23,66.2,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.5 model card: claude-opus-4.5=66.2.
gemini-3-pro,finsearchcompt23,49.9,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.5 model card: gemini-3-pro=49.9.
deepseek-v3.2,finsearchcompt23,59.1,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.5 model card: deepseek-v3.2=59.1.
kimi-k2.5,seal_0,57.4,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.5 model card: kimi-k2.5=57.4.
gpt-5.2,seal_0,45.0,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.5 model card: gpt-5.2=45.0.
claude-opus-4.5,seal_0,47.7,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.5 model card: claude-opus-4.5=47.7.
gemini-3-pro,seal_0,45.5,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.5 model card: gemini-3-pro=45.5.
deepseek-v3.2,seal_0,49.5,https://huggingface.co/moonshotai/Kimi-K2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",True,model_card,verified,Kimi K2.5 model card: deepseek-v3.2=49.5.
gpt-5.2,mmlu_redux,95.0,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,model_card,verified,"Qwen3.5-397B-A17B HF model card cross-model table (Qwen self-test of 3rd-party model, mc=false): mmlu_redux 95.0 (3rd-party Qwen self-test)."
claude-opus-4.5,mmlu_redux,95.6,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,model_card,verified,"Qwen3.5-397B-A17B HF model card cross-model table (Qwen self-test of 3rd-party model, mc=false): mmlu_redux 95.6 (3rd-party Qwen self-test)."
gemini-3-pro,mmlu_redux,95.9,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,model_card,verified,"Qwen3.5-397B-A17B HF model card cross-model table (Qwen self-test of 3rd-party model, mc=false): mmlu_redux 95.9 (3rd-party Qwen self-test)."
kimi-k2.5,mmlu_redux,94.5,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,model_card,verified,"Qwen3.5-397B-A17B HF model card cross-model table (Qwen self-test of 3rd-party model, mc=false): mmlu_redux 94.5 (3rd-party Qwen self-test)."
gpt-5.2,supergpqa,67.9,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,model_card,verified,"Qwen3.5-397B-A17B HF model card cross-model table (Qwen self-test of 3rd-party model, mc=false): supergpqa 67.9 (3rd-party Qwen self-test)."
claude-opus-4.5,supergpqa,70.6,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,model_card,verified,"Qwen3.5-397B-A17B HF model card cross-model table (Qwen self-test of 3rd-party model, mc=false): supergpqa 70.6 (3rd-party Qwen self-test)."
gemini-3-pro,supergpqa,74.0,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,model_card,verified,"Qwen3.5-397B-A17B HF model card cross-model table (Qwen self-test of 3rd-party model, mc=false): supergpqa 74.0 (3rd-party Qwen self-test)."
kimi-k2.5,supergpqa,69.2,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,model_card,verified,"Qwen3.5-397B-A17B HF model card cross-model table (Qwen self-test of 3rd-party model, mc=false): supergpqa 69.2 (3rd-party Qwen self-test)."
gpt-5.2,ifbench,75.4,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,model_card,verified,"Qwen3.5-397B-A17B HF model card cross-model table (Qwen self-test of 3rd-party model, mc=false): ifbench 75.4 (3rd-party Qwen self-test)."
claude-opus-4.5,ifbench,58.0,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,model_card,verified,"Qwen3.5-397B-A17B HF model card cross-model table (Qwen self-test of 3rd-party model, mc=false): ifbench 58.0 (3rd-party Qwen self-test)."
gemini-3-pro,ifbench,70.4,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,model_card,verified,"Qwen3.5-397B-A17B HF model card cross-model table (Qwen self-test of 3rd-party model, mc=false): ifbench 70.4 (3rd-party Qwen self-test)."
kimi-k2.5,ifbench,70.2,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,model_card,verified,"Qwen3.5-397B-A17B HF model card cross-model table (Qwen self-test of 3rd-party model, mc=false): ifbench 70.2 (3rd-party Qwen self-test)."
gpt-5.2,multichallenge,57.9,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,model_card,verified,"Qwen3.5-397B-A17B HF model card cross-model table (Qwen self-test of 3rd-party model, mc=false): multichallenge 57.9 (3rd-party Qwen self-test)."
claude-opus-4.5,multichallenge,54.2,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,model_card,verified,"Qwen3.5-397B-A17B HF model card cross-model table (Qwen self-test of 3rd-party model, mc=false): multichallenge 54.2 (3rd-party Qwen self-test)."
gemini-3-pro,multichallenge,64.2,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,model_card,verified,"Qwen3.5-397B-A17B HF model card cross-model table (Qwen self-test of 3rd-party model, mc=false): multichallenge 64.2 (3rd-party Qwen self-test)."
kimi-k2.5,multichallenge,62.7,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,model_card,verified,"Qwen3.5-397B-A17B HF model card cross-model table (Qwen self-test of 3rd-party model, mc=false): multichallenge 62.7 (3rd-party Qwen self-test)."
gpt-5.2,aime_2026,96.7,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,model_card,verified,"Qwen3.5-397B-A17B HF model card cross-model table (Qwen self-test of 3rd-party model, mc=false): aime_2026 96.7 (3rd-party Qwen self-test)."
claude-opus-4.5,aime_2026,93.3,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,model_card,verified,"Qwen3.5-397B-A17B HF model card cross-model table (Qwen self-test of 3rd-party model, mc=false): aime_2026 93.3 (3rd-party Qwen self-test)."
gemini-3-pro,aime_2026,90.6,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,model_card,verified,"Qwen3.5-397B-A17B HF model card cross-model table (Qwen self-test of 3rd-party model, mc=false): aime_2026 90.6 (3rd-party Qwen self-test)."
kimi-k2.5,mmmlu,86.0,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,model_card,verified,"Qwen3.5-397B-A17B HF model card cross-model table (Qwen self-test of 3rd-party model, mc=false): mmmlu 86.0 (3rd-party Qwen self-test)."
claude-opus-4.5,global_piqa,91.6,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,model_card,verified,"Qwen3.5-397B-A17B HF model card cross-model table (Qwen self-test of 3rd-party model, mc=false): global_piqa 91.6 (3rd-party Qwen self-test)."
kimi-k2.5,global_piqa,89.3,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,model_card,verified,"Qwen3.5-397B-A17B HF model card cross-model table (Qwen self-test of 3rd-party model, mc=false): global_piqa 89.3 (3rd-party Qwen self-test)."
gpt-5.2,mcpmark,57.5,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,model_card,verified,"Qwen3.5-397B-A17B HF model card cross-model table (Qwen self-test of 3rd-party model, mc=false): mcpmark 57.5 (3rd-party Qwen self-test)."
claude-opus-4.5,mcpmark,42.3,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,model_card,verified,"Qwen3.5-397B-A17B HF model card cross-model table (Qwen self-test of 3rd-party model, mc=false): mcpmark 42.3 (3rd-party Qwen self-test)."
gemini-3-pro,mcpmark,53.9,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,model_card,verified,"Qwen3.5-397B-A17B HF model card cross-model table (Qwen self-test of 3rd-party model, mc=false): mcpmark 53.9 (3rd-party Qwen self-test)."
claude-opus-4.5,screenspot_pro,45.7,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,model_card,verified,"Qwen3.5-397B-A17B HF model card cross-model table (Qwen self-test of 3rd-party model, mc=false): screenspot_pro 45.7 (3rd-party Qwen self-test)."
claude-opus-4.5,video_mme,77.6,https://huggingface.co/Qwen/Qwen3.5-397B-A17B,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,model_card,verified,"Qwen3.5-397B-A17B HF model card cross-model table (Qwen self-test of 3rd-party model, mc=false): video_mme 77.6 (3rd-party Qwen self-test)."
kimi-k2,hmmt_feb_2025,38.8,https://huggingface.co/moonshotai/Kimi-K2-Thinking,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2-Thinking model card: kimi-k2=38.8.
grok-4,hmmt_feb_2025,90.0,https://huggingface.co/moonshotai/Kimi-K2-Thinking,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2-Thinking model card: grok-4=90.0.
kimi-k2,imo_answerbench,45.8,https://huggingface.co/moonshotai/Kimi-K2-Thinking,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2-Thinking model card: kimi-k2=45.8.
grok-4,imo_answerbench,73.1,https://huggingface.co/moonshotai/Kimi-K2-Thinking,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2-Thinking model card: grok-4=73.1.
gpt-5,mmlu_redux,95.3,https://huggingface.co/moonshotai/Kimi-K2-Thinking,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2-Thinking model card: gpt-5=95.3.
claude-sonnet-4.5,mmlu_redux,95.6,https://huggingface.co/moonshotai/Kimi-K2-Thinking,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2-Thinking model card: claude-sonnet-4.5=95.6.
deepseek-v3.2,mmlu_redux,93.7,https://huggingface.co/moonshotai/Kimi-K2-Thinking,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2-Thinking model card: deepseek-v3.2=93.7.
kimi-k2-thinking,longform_writing,73.8,https://huggingface.co/moonshotai/Kimi-K2-Thinking,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2-Thinking model card: kimi-k2-thinking=73.8.
gpt-5,longform_writing,71.4,https://huggingface.co/moonshotai/Kimi-K2-Thinking,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2-Thinking model card: gpt-5=71.4.
claude-sonnet-4.5,longform_writing,79.8,https://huggingface.co/moonshotai/Kimi-K2-Thinking,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2-Thinking model card: claude-sonnet-4.5=79.8.
kimi-k2,longform_writing,62.8,https://huggingface.co/moonshotai/Kimi-K2-Thinking,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2-Thinking model card: kimi-k2=62.8.
deepseek-v3.2,longform_writing,72.5,https://huggingface.co/moonshotai/Kimi-K2-Thinking,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2-Thinking model card: deepseek-v3.2=72.5.
kimi-k2-thinking,healthbench,58.0,https://huggingface.co/moonshotai/Kimi-K2-Thinking,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2-Thinking model card: kimi-k2-thinking=58.0.
gpt-5,healthbench,67.2,https://huggingface.co/moonshotai/Kimi-K2-Thinking,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2-Thinking model card: gpt-5=67.2.
claude-sonnet-4.5,healthbench,44.2,https://huggingface.co/moonshotai/Kimi-K2-Thinking,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2-Thinking model card: claude-sonnet-4.5=44.2.
kimi-k2,healthbench,43.8,https://huggingface.co/moonshotai/Kimi-K2-Thinking,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2-Thinking model card: kimi-k2=43.8.
deepseek-v3.2,healthbench,46.9,https://huggingface.co/moonshotai/Kimi-K2-Thinking,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2-Thinking model card: deepseek-v3.2=46.9.
gpt-5,scicode,42.9,https://huggingface.co/moonshotai/Kimi-K2-Thinking,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2-Thinking model card: gpt-5=42.9.
claude-sonnet-4.5,scicode,44.7,https://huggingface.co/moonshotai/Kimi-K2-Thinking,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2-Thinking model card: claude-sonnet-4.5=44.7.
kimi-k2,scicode,30.7,https://huggingface.co/moonshotai/Kimi-K2-Thinking,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2-Thinking model card: kimi-k2=30.7.
kimi-k2,livecodebench_v6,56.1,https://huggingface.co/moonshotai/Kimi-K2-Thinking,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2-Thinking model card: kimi-k2=56.1.
gpt-5,ojbench,56.2,https://huggingface.co/moonshotai/Kimi-K2-Thinking,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2-Thinking model card: gpt-5=56.2.
claude-sonnet-4.5,ojbench,30.4,https://huggingface.co/moonshotai/Kimi-K2-Thinking,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2-Thinking model card: claude-sonnet-4.5=30.4.
kimi-k2,ojbench,25.5,https://huggingface.co/moonshotai/Kimi-K2-Thinking,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none""}",True,model_card,verified,Kimi K2-Thinking model card: kimi-k2=25.5.
gpt-oss-120b,aider_polyglot_diff,44.4,https://arxiv.org/abs/2508.10925,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""harmony"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""with tools""}",True,tech_report,verified,gpt-oss model card Table 3 (high effort): Aider Polyglot 44.4 (high)
gpt-oss-120b,mmmlu,81.3,https://arxiv.org/abs/2508.10925,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""harmony"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""with tools""}",True,tech_report,verified,gpt-oss model card Table 3 (high effort): MMMLU Average 81.3
gpt-oss-120b,healthbench,57.6,https://arxiv.org/abs/2508.10925,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""harmony"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""with tools""}",True,tech_report,verified,gpt-oss model card Table 3 (high effort): HealthBench 57.6
gpt-oss-120b,healthbench_hard,30.0,https://arxiv.org/abs/2508.10925,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""harmony"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""with tools""}",True,tech_report,verified,gpt-oss model card Table 3 (high effort): HealthBench Hard 30.0
gpt-oss-120b,healthbench_consensus,89.9,https://arxiv.org/abs/2508.10925,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""harmony"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""with tools""}",True,tech_report,verified,gpt-oss model card Table 3 (high effort): HealthBench Consensus 89.9
gpt-oss-120b,hle,19.0,https://arxiv.org/abs/2508.10925,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""harmony"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""with tools""}",True,tech_report,verified,gpt-oss model card Table 3 (high effort): HLE with-tools 19.0
gpt-oss-120b,tau_bench_airline,49.2,https://arxiv.org/abs/2508.10925,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""harmony"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""with tools""}",True,tech_report,verified,gpt-oss model card Table 3 (high effort): Tau-Bench Airline with-tools 49.2
gpt-oss-20b,aider_polyglot_diff,34.2,https://arxiv.org/abs/2508.10925,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""harmony"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""with tools""}",True,tech_report,verified,gpt-oss model card Table 3 (high effort): Aider Polyglot 34.2 (high)
gpt-oss-20b,mmmlu,75.7,https://arxiv.org/abs/2508.10925,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""harmony"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""with tools""}",True,tech_report,verified,gpt-oss model card Table 3 (high effort): MMMLU Average 75.7
gpt-oss-20b,healthbench,42.5,https://arxiv.org/abs/2508.10925,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""harmony"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""with tools""}",True,tech_report,verified,gpt-oss model card Table 3 (high effort): HealthBench 42.5
gpt-oss-20b,healthbench_hard,10.8,https://arxiv.org/abs/2508.10925,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""harmony"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""with tools""}",True,tech_report,verified,gpt-oss model card Table 3 (high effort): HealthBench Hard 10.8
gpt-oss-20b,healthbench_consensus,82.6,https://arxiv.org/abs/2508.10925,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""harmony"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""with tools""}",True,tech_report,verified,gpt-oss model card Table 3 (high effort): HealthBench Consensus 82.6
gpt-oss-20b,hle,17.3,https://arxiv.org/abs/2508.10925,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""harmony"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""with tools""}",True,tech_report,verified,gpt-oss model card Table 3 (high effort): HLE with-tools 17.3
gpt-oss-20b,tau_bench_retail,54.8,https://arxiv.org/abs/2508.10925,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""harmony"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""with tools""}",True,tech_report,verified,gpt-oss model card Table 3 (high effort): Tau-Bench Retail with-tools 54.8
gpt-oss-20b,tau_bench_airline,38.0,https://arxiv.org/abs/2508.10925,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""harmony"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""with tools""}",True,tech_report,verified,gpt-oss model card Table 3 (high effort): Tau-Bench Airline with-tools 38.0
kimi-k2,aider_polyglot_diff,60.0,https://huggingface.co/moonshotai/Kimi-K2-Instruct,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,model_card,verified,Kimi K2-Instruct model card: 60.0.
kimi-k2,cnmo_2024,74.3,https://huggingface.co/moonshotai/Kimi-K2-Instruct,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Kimi K2-Instruct model card: 74.3.
gemma-3-27b,bird_sql,54.4,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report Table 6 IT 27B: Bird-SQL Dev 54.4
gemma-3-27b,facts_grounding,74.9,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report Table 6 IT 27B: FACTS Grounding 74.9
gemma-3-27b,global_mmlu_lite,75.1,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report Table 6 IT 27B: Global MMLU-Lite 75.1
gemma-3-27b,math,89.0,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report Table 6 IT 27B: MATH 89.0
gemma-3-27b,hiddenmath,60.3,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report Table 6 IT 27B: HiddenMath 60.3
gemma-3-27b,mmmu,64.9,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report Table 6 IT 27B: MMMU 64.9
gemma-4-31b,bigbench_extra_hard,74.4,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-26b-a4b,bigbench_extra_hard,64.8,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-e4b,bigbench_extra_hard,33.1,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-e2b,bigbench_extra_hard,21.9,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-31b,medxpertqa_mm,61.3,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none"",""visual_tokens"":1120}",True,tech_report,verified,"Gemma 4 Technical Report Table 6; maximum supported resolution, 1,120 visual tokens."
gemma-4-26b-a4b,medxpertqa_mm,58.1,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none"",""visual_tokens"":1120}",True,tech_report,verified,"Gemma 4 Technical Report Table 6; maximum supported resolution, 1,120 visual tokens."
gemma-4-e4b,medxpertqa_mm,28.7,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none"",""visual_tokens"":1120}",True,tech_report,verified,"Gemma 4 Technical Report Table 6; maximum supported resolution, 1,120 visual tokens."
gemma-4-e2b,medxpertqa_mm,23.5,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none"",""visual_tokens"":1120}",True,tech_report,verified,"Gemma 4 Technical Report Table 6; maximum supported resolution, 1,120 visual tokens."
gemma-4-e4b,fleurs,0.075,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-4-e2b,fleurs,0.09,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
minimax-m2.5,aime_2025,86.3,https://huggingface.co/MiniMaxAI/MiniMax-M2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2.5 model card: minimax-m2.5=86.3.
minimax-m2.1,aime_2025,83.0,https://huggingface.co/MiniMaxAI/MiniMax-M2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2.5 model card: minimax-m2.1=83.0.
minimax-m2.5,gpqa_diamond,85.2,https://huggingface.co/MiniMaxAI/MiniMax-M2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2.5 model card: minimax-m2.5=85.2.
minimax-m2.1,gpqa_diamond,83.0,https://huggingface.co/MiniMaxAI/MiniMax-M2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2.5 model card: minimax-m2.1=83.0.
minimax-m2.5,hle,19.4,https://huggingface.co/MiniMaxAI/MiniMax-M2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2.5 model card: minimax-m2.5=19.4.
minimax-m2.1,hle,22.2,https://huggingface.co/MiniMaxAI/MiniMax-M2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2.5 model card: minimax-m2.1=22.2.
minimax-m2.5,scicode,44.4,https://huggingface.co/MiniMaxAI/MiniMax-M2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2.5 model card: minimax-m2.5=44.4.
minimax-m2.1,scicode,41.0,https://huggingface.co/MiniMaxAI/MiniMax-M2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2.5 model card: minimax-m2.1=41.0.
minimax-m2.5,ifbench,70.0,https://huggingface.co/MiniMaxAI/MiniMax-M2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2.5 model card: minimax-m2.5=70.0.
minimax-m2.1,ifbench,70.0,https://huggingface.co/MiniMaxAI/MiniMax-M2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2.5 model card: minimax-m2.1=70.0.
claude-sonnet-4.5,ifbench,57.0,https://huggingface.co/MiniMaxAI/MiniMax-M2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2.5 model card: claude-sonnet-4.5=57.0.
claude-opus-4.6,ifbench,53.0,https://huggingface.co/MiniMaxAI/MiniMax-M2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2.5 model card: claude-opus-4.6=53.0.
minimax-m2.5,aa_lcr,69.5,https://huggingface.co/MiniMaxAI/MiniMax-M2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2.5 model card: minimax-m2.5=69.5.
minimax-m2.1,aa_lcr,62.0,https://huggingface.co/MiniMaxAI/MiniMax-M2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2.5 model card: minimax-m2.1=62.0.
claude-sonnet-4.5,aa_lcr,66.0,https://huggingface.co/MiniMaxAI/MiniMax-M2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2.5 model card: claude-sonnet-4.5=66.0.
claude-opus-4.6,aa_lcr,71.0,https://huggingface.co/MiniMaxAI/MiniMax-M2.5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2.5 model card: claude-opus-4.6=71.0.
minimax-m2,multi_swe_bench,36.2,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,model_card,verified,MiniMax M2 model card: minimax-m2=36.2.
claude-sonnet-4,multi_swe_bench,35.7,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,model_card,verified,MiniMax M2 model card: claude-sonnet-4=35.7.
claude-sonnet-4.5,multi_swe_bench,44.3,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,model_card,verified,MiniMax M2 model card: claude-sonnet-4.5=44.3.
glm-4.6,multi_swe_bench,30.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,model_card,verified,MiniMax M2 model card: glm-4.6=30.
kimi-k2,multi_swe_bench,33.5,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,model_card,verified,MiniMax M2 model card: kimi-k2=33.5.
deepseek-v3.2,multi_swe_bench,30.6,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,model_card,verified,MiniMax M2 model card: deepseek-v3.2=30.6.
claude-sonnet-4,swe_bench_multilingual,56.9,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,model_card,verified,MiniMax M2 model card: claude-sonnet-4=56.9.
minimax-m2,artifactsbench,66.8,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: minimax-m2=66.8.
claude-sonnet-4,artifactsbench,57.3,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: claude-sonnet-4=57.3.
claude-sonnet-4.5,artifactsbench,61.5,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: claude-sonnet-4.5=61.5.
gemini-2.5-pro,artifactsbench,57.7,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: gemini-2.5-pro=57.7.
gpt-5,artifactsbench,73.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: gpt-5=73.
glm-4.6,artifactsbench,59.8,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: glm-4.6=59.8.
kimi-k2,artifactsbench,54.2,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: kimi-k2=54.2.
deepseek-v3.2,artifactsbench,55.8,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: deepseek-v3.2=55.8.
claude-sonnet-4,browsecomp,12.2,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: claude-sonnet-4=12.2.
gemini-2.5-pro,browsecomp,9.9,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: gemini-2.5-pro=9.9.
kimi-k2,browsecomp,14.1,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: kimi-k2=14.1.
minimax-m2,browsecomp_zh,48.5,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: minimax-m2=48.5.
claude-sonnet-4,browsecomp_zh,29.1,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: claude-sonnet-4=29.1.
gemini-2.5-pro,browsecomp_zh,32.2,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: gemini-2.5-pro=32.2.
kimi-k2,browsecomp_zh,28.8,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: kimi-k2=28.8.
minimax-m2,gaia,75.7,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: minimax-m2=75.7.
claude-sonnet-4,gaia,68.3,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: claude-sonnet-4=68.3.
claude-sonnet-4.5,gaia,71.2,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: claude-sonnet-4.5=71.2.
gemini-2.5-pro,gaia,60.2,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: gemini-2.5-pro=60.2.
gpt-5,gaia,76.4,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: gpt-5=76.4.
glm-4.6,gaia,71.9,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: glm-4.6=71.9.
kimi-k2,gaia,60.2,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: kimi-k2=60.2.
deepseek-v3.2,gaia,63.5,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: deepseek-v3.2=63.5.
minimax-m2,xbench_deepsearch,72.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: minimax-m2=72.
claude-sonnet-4,xbench_deepsearch,64.6,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: claude-sonnet-4=64.6.
claude-sonnet-4.5,xbench_deepsearch,66.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: claude-sonnet-4.5=66.
gemini-2.5-pro,xbench_deepsearch,56.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: gemini-2.5-pro=56.
gpt-5,xbench_deepsearch,77.8,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: gpt-5=77.8.
glm-4.6,xbench_deepsearch,70.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: glm-4.6=70.
kimi-k2,xbench_deepsearch,61.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: kimi-k2=61.
deepseek-v3.2,xbench_deepsearch,71.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: deepseek-v3.2=71.
minimax-m2,tau2_bench_avg,77.2,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: minimax-m2=77.2.
claude-sonnet-4,tau2_bench_avg,65.5,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: claude-sonnet-4=65.5.
claude-sonnet-4.5,tau2_bench_avg,84.7,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: claude-sonnet-4.5=84.7.
gemini-2.5-pro,tau2_bench_avg,59.2,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: gemini-2.5-pro=59.2.
gpt-5,tau2_bench_avg,80.1,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: gpt-5=80.1.
glm-4.6,tau2_bench_avg,75.9,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: glm-4.6=75.9.
kimi-k2,tau2_bench_avg,70.3,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: kimi-k2=70.3.
deepseek-v3.2,tau2_bench_avg,66.7,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: deepseek-v3.2=66.7.
minimax-m2,finsearchcomp_global,65.5,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: minimax-m2=65.5.
claude-sonnet-4,finsearchcomp_global,42.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: claude-sonnet-4=42.
claude-sonnet-4.5,finsearchcomp_global,60.8,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: claude-sonnet-4.5=60.8.
gemini-2.5-pro,finsearchcomp_global,42.6,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: gemini-2.5-pro=42.6.
gpt-5,finsearchcomp_global,63.9,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: gpt-5=63.9.
glm-4.6,finsearchcomp_global,29.2,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: glm-4.6=29.2.
kimi-k2,finsearchcomp_global,29.5,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: kimi-k2=29.5.
deepseek-v3.2,finsearchcomp_global,26.2,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: deepseek-v3.2=26.2.
minimax-m2,agentcompany,36.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: minimax-m2=36.
claude-sonnet-4,agentcompany,37.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: claude-sonnet-4=37.
claude-sonnet-4.5,agentcompany,41.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: claude-sonnet-4.5=41.
gemini-2.5-pro,agentcompany,39.3,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: gemini-2.5-pro=39.3.
glm-4.6,agentcompany,35.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: glm-4.6=35.
kimi-k2,agentcompany,30.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: kimi-k2=30.
deepseek-v3.2,agentcompany,34.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: deepseek-v3.2=34.
minimax-m2,scicode,36.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2 model card: minimax-m2=36.
claude-sonnet-4,scicode,40.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2 model card: claude-sonnet-4=40.
gemini-2.5-pro,scicode,43.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2 model card: gemini-2.5-pro=43.
glm-4.6,scicode,38.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2 model card: glm-4.6=38.
minimax-m2,ifbench,72.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2 model card: minimax-m2=72.
gpt-5,ifbench,73.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2 model card: gpt-5=73.
glm-4.6,ifbench,43.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2 model card: glm-4.6=43.
kimi-k2,ifbench,42.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2 model card: kimi-k2=42.
deepseek-v3.2,ifbench,54.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2 model card: deepseek-v3.2=54.
minimax-m2,aa_lcr,61.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2 model card: minimax-m2=61.
claude-sonnet-4,aa_lcr,65.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2 model card: claude-sonnet-4=65.
gemini-2.5-pro,aa_lcr,66.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2 model card: gemini-2.5-pro=66.
glm-4.6,aa_lcr,54.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2 model card: glm-4.6=54.
kimi-k2,aa_lcr,52.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2 model card: kimi-k2=52.
minimax-m2,tau2_bench_telecom,87.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: minimax-m2=87.
gemini-2.5-pro,tau2_bench_telecom,54.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: gemini-2.5-pro=54.
glm-4.6,tau2_bench_telecom,71.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: glm-4.6=71.
minimax-m2,terminal_bench_hard,24.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: minimax-m2=24.
claude-sonnet-4,terminal_bench_hard,30.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: claude-sonnet-4=30.
gemini-2.5-pro,terminal_bench_hard,25.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: gemini-2.5-pro=25.
kimi-k2,terminal_bench_hard,23.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2 model card: kimi-k2=23.
claude-sonnet-4,aa_intelligence_index,57.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2 model card: claude-sonnet-4=57.
claude-sonnet-4.5,aa_intelligence_index,63.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2 model card: claude-sonnet-4.5=63.
glm-4.6,aa_intelligence_index,56.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2 model card: glm-4.6=56.
kimi-k2,aa_intelligence_index,50.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2 model card: kimi-k2=50.
deepseek-v3.2,aa_intelligence_index,57.0,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,MiniMax M2 model card: deepseek-v3.2=57.
gemini-2.0-pro,mmlu_pro,79.1,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: mmlu_pro 79.1 (gemini-2.0-pro column).
gemini-2.0-pro,livecodebench,36.0,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: livecodebench 36.0 (gemini-2.0-pro column).
gemini-2.0-pro,bird_sql,59.3,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: bird_sql 59.3 (gemini-2.0-pro column).
gemini-2.0-pro,gpqa_diamond,64.7,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: gpqa_diamond 64.7 (gemini-2.0-pro column).
gemini-2.0-pro,simpleqa,44.3,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: simpleqa 44.3 (gemini-2.0-pro column).
gemini-2.0-pro,facts_grounding,82.8,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: facts_grounding 82.8 (gemini-2.0-pro column).
gemini-2.0-pro,global_mmlu_lite,86.5,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: global_mmlu_lite 86.5 (gemini-2.0-pro column).
gemini-2.0-pro,math,91.8,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: math 91.8 (gemini-2.0-pro column).
gemini-2.0-pro,hiddenmath,65.2,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: hiddenmath 65.2 (gemini-2.0-pro column).
gemini-2.0-pro,mmmu,72.7,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: mmmu 72.7 (gemini-2.0-pro column).
gemma-2-2b,mmlu_pro,15.6,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: mmlu_pro 15.6 (gemma-2-2b column).
gemma-2-2b,livecodebench,1.2,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: livecodebench 1.2 (gemma-2-2b column).
gemma-2-2b,bird_sql,12.2,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: bird_sql 12.2 (gemma-2-2b column).
gemma-2-2b,gpqa_diamond,24.7,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: gpqa_diamond 24.7 (gemma-2-2b column).
gemma-2-2b,simpleqa,2.8,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: simpleqa 2.8 (gemma-2-2b column).
gemma-2-2b,facts_grounding,43.8,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: facts_grounding 43.8 (gemma-2-2b column).
gemma-2-2b,global_mmlu_lite,41.9,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: global_mmlu_lite 41.9 (gemma-2-2b column).
gemma-2-2b,math,27.2,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: math 27.2 (gemma-2-2b column).
gemma-2-2b,hiddenmath,1.8,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: hiddenmath 1.8 (gemma-2-2b column).
gemma-2-9b,mmlu_pro,46.8,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: mmlu_pro 46.8 (gemma-2-9b column).
gemma-2-9b,livecodebench,10.8,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: livecodebench 10.8 (gemma-2-9b column).
gemma-2-9b,bird_sql,33.8,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: bird_sql 33.8 (gemma-2-9b column).
gemma-2-9b,gpqa_diamond,28.8,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: gpqa_diamond 28.8 (gemma-2-9b column).
gemma-2-9b,simpleqa,5.3,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: simpleqa 5.3 (gemma-2-9b column).
gemma-2-9b,facts_grounding,62.0,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: facts_grounding 62.0 (gemma-2-9b column).
gemma-2-9b,global_mmlu_lite,64.8,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: global_mmlu_lite 64.8 (gemma-2-9b column).
gemma-2-9b,math,49.4,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: math 49.4 (gemma-2-9b column).
gemma-2-9b,hiddenmath,10.4,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: hiddenmath 10.4 (gemma-2-9b column).
gemma-2-27b,mmlu_pro,56.9,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: mmlu_pro 56.9 (gemma-2-27b column).
gemma-2-27b,livecodebench,20.4,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: livecodebench 20.4 (gemma-2-27b column).
gemma-2-27b,bird_sql,46.7,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: bird_sql 46.7 (gemma-2-27b column).
gemma-2-27b,gpqa_diamond,34.3,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: gpqa_diamond 34.3 (gemma-2-27b column).
gemma-2-27b,simpleqa,9.2,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: simpleqa 9.2 (gemma-2-27b column).
gemma-2-27b,facts_grounding,62.4,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: facts_grounding 62.4 (gemma-2-27b column).
gemma-2-27b,global_mmlu_lite,68.6,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: global_mmlu_lite 68.6 (gemma-2-27b column).
gemma-2-27b,math,55.6,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: math 55.6 (gemma-2-27b column).
gemma-2-27b,hiddenmath,14.8,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: hiddenmath 14.8 (gemma-2-27b column).
gemma-3-1b,mmlu_pro,14.7,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: mmlu_pro 14.7 (gemma-3-1b column).
gemma-3-1b,livecodebench,1.9,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: livecodebench 1.9 (gemma-3-1b column).
gemma-3-1b,bird_sql,6.4,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: bird_sql 6.4 (gemma-3-1b column).
gemma-3-1b,gpqa_diamond,19.2,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: gpqa_diamond 19.2 (gemma-3-1b column).
gemma-3-1b,simpleqa,2.2,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: simpleqa 2.2 (gemma-3-1b column).
gemma-3-1b,facts_grounding,36.4,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: facts_grounding 36.4 (gemma-3-1b column).
gemma-3-1b,global_mmlu_lite,34.2,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: global_mmlu_lite 34.2 (gemma-3-1b column).
gemma-3-1b,math,48.0,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: math 48.0 (gemma-3-1b column).
gemma-3-1b,hiddenmath,15.8,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: hiddenmath 15.8 (gemma-3-1b column).
gemma-3-4b,mmlu_pro,43.6,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: mmlu_pro 43.6 (gemma-3-4b column).
gemma-3-4b,livecodebench,12.6,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: livecodebench 12.6 (gemma-3-4b column).
gemma-3-4b,bird_sql,36.3,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: bird_sql 36.3 (gemma-3-4b column).
gemma-3-4b,gpqa_diamond,30.8,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: gpqa_diamond 30.8 (gemma-3-4b column).
gemma-3-4b,simpleqa,4.0,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: simpleqa 4.0 (gemma-3-4b column).
gemma-3-4b,facts_grounding,70.1,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: facts_grounding 70.1 (gemma-3-4b column).
gemma-3-4b,global_mmlu_lite,54.5,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: global_mmlu_lite 54.5 (gemma-3-4b column).
gemma-3-4b,math,75.6,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: math 75.6 (gemma-3-4b column).
gemma-3-4b,hiddenmath,43.0,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: hiddenmath 43.0 (gemma-3-4b column).
gemma-3-4b,mmmu,48.8,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: mmmu 48.8 (gemma-3-4b column).
gemma-3-12b,mmlu_pro,60.6,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: mmlu_pro 60.6 (gemma-3-12b column).
gemma-3-12b,livecodebench,24.6,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: livecodebench 24.6 (gemma-3-12b column).
gemma-3-12b,bird_sql,47.9,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: bird_sql 47.9 (gemma-3-12b column).
gemma-3-12b,gpqa_diamond,40.9,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: gpqa_diamond 40.9 (gemma-3-12b column).
gemma-3-12b,simpleqa,6.3,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: simpleqa 6.3 (gemma-3-12b column).
gemma-3-12b,facts_grounding,75.8,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: facts_grounding 75.8 (gemma-3-12b column).
gemma-3-12b,global_mmlu_lite,69.5,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: global_mmlu_lite 69.5 (gemma-3-12b column).
gemma-3-12b,math,83.8,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: math 83.8 (gemma-3-12b column).
gemma-3-12b,hiddenmath,54.5,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: hiddenmath 54.5 (gemma-3-12b column).
gemma-3-12b,mmmu,59.6,https://arxiv.org/abs/2503.19786,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Gemma 3 tech report (arxiv:2503.19786) Table 6 IT: mmmu 59.6 (gemma-3-12b column).
seed-thinking-v1.5,beyond_aime,48.0,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: seed-thinking-v1.5=48.0.
deepseek-r1,beyond_aime,42.4,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: deepseek-r1=42.4.
o3-mini-high,beyond_aime,63.6,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: o3-mini-high=63.6.
gemini-2.5-pro,beyond_aime,58.8,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: gemini-2.5-pro=58.8.
seed-thinking-v1.5,supergpqa,62.1,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: seed-thinking-v1.5=62.1.
deepseek-r1,supergpqa,60.5,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: deepseek-r1=60.5.
o3-mini-high,supergpqa,52.2,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: o3-mini-high=52.2.
grok-3-beta,supergpqa,62.8,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: grok-3-beta=62.8.
gemini-2.5-pro,supergpqa,65.3,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: gemini-2.5-pro=65.3.
seed-thinking-v1.5,codeforces_pass8,55.0,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: seed-thinking-v1.5=55.0.
deepseek-r1,codeforces_pass8,45.0,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: deepseek-r1=45.0.
o3-mini-high,codeforces_pass8,67.5,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: o3-mini-high=67.5.
gemini-2.5-pro,codeforces_pass8,56.3,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: gemini-2.5-pro=56.3.
seed-thinking-v1.5,codeforces_avg8,36.3,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: seed-thinking-v1.5=36.3.
deepseek-r1,codeforces_avg8,32.0,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: deepseek-r1=32.0.
o3-mini-high,codeforces_avg8,50.9,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: o3-mini-high=50.9.
gemini-2.5-pro,codeforces_avg8,40.3,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: gemini-2.5-pro=40.3.
seed-thinking-v1.5,livecodebench_v5,64.9,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: seed-thinking-v1.5=64.9.
deepseek-r1,livecodebench_v5,64.3,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: deepseek-r1=64.3.
o3-mini-high,livecodebench_v5,74.1,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: o3-mini-high=74.1.
grok-3-beta,livecodebench_v5,70.6,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: grok-3-beta=70.6.
gemini-2.5-pro,livecodebench_v5,70.4,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: gemini-2.5-pro=70.4.
seed-thinking-v1.5,aider_polyglot_diff,54.2,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: seed-thinking-v1.5=54.2.
seed-thinking-v1.5,arc_agi_1,39.9,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: seed-thinking-v1.5=39.9.
seed-thinking-v1.5,collie,73.1,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: seed-thinking-v1.5=73.1.
deepseek-r1,collie,34.2,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: deepseek-r1=34.2.
grok-3-beta,collie,33.6,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: grok-3-beta=33.6.
gemini-2.5-pro,collie,62.5,https://arxiv.org/abs/2504.13914,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Seed-Thinking-v1.5 paper Table 2: gemini-2.5-pro=62.5.
claude-opus-4.5,simpleqa_verified,48.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-opus-4.5=48.6.
doubao-seed-2.0-pro,simpleqa_verified,36.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=36.0.
gpt-5.2,healthbench,63.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gpt-5.2=63.3.
claude-opus-4.5,healthbench,36.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-opus-4.5=36.3.
gemini-3-pro,healthbench,37.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gemini-3-pro=37.9.
doubao-seed-2.0-pro,healthbench,57.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=57.7.
gpt-5.2,healthbench_hard,42.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gpt-5.2=42.0.
claude-sonnet-4.5,healthbench_hard,10.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-sonnet-4.5=10.9.
claude-opus-4.5,healthbench_hard,11.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-opus-4.5=11.0.
gemini-3-pro,healthbench_hard,15.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gemini-3-pro=15.0.
doubao-seed-2.0-pro,healthbench_hard,29.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=29.1.
claude-sonnet-4.5,supergpqa,65.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-sonnet-4.5=65.5.
doubao-seed-2.0-pro,supergpqa,68.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=68.7.
gpt-5.2,lpfqa,54.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gpt-5.2=54.4.
claude-sonnet-4.5,lpfqa,54.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-sonnet-4.5=54.9.
claude-opus-4.5,lpfqa,52.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-opus-4.5=52.6.
gemini-3-pro,lpfqa,51.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gemini-3-pro=51.2.
doubao-seed-2.0-pro,lpfqa,52.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=52.6.
gpt-5.2,encyclo_k,61.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gpt-5.2=61.0.
claude-sonnet-4.5,encyclo_k,58.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-sonnet-4.5=58.0.
claude-opus-4.5,encyclo_k,63.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-opus-4.5=63.3.
gemini-3-pro,encyclo_k,64.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gemini-3-pro=64.9.
doubao-seed-2.0-pro,encyclo_k,65.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=65.7.
claude-sonnet-4.5,aime_2026,82.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-sonnet-4.5=82.5.
doubao-seed-2.0-pro,aime_2026,94.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=94.2.
doubao-seed-2.0-pro,hmmt_feb_2025,97.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=97.3.
doubao-seed-2.0-pro,hmmt_nov_2025,93.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=93.3.
claude-opus-4.5,matharena_apex_2025,1.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-opus-4.5=1.6.
doubao-seed-2.0-pro,matharena_apex_2025,20.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=20.3.
gpt-5.2,apex_shortlist,80.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gpt-5.2=80.1.
claude-sonnet-4.5,apex_shortlist,26.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-sonnet-4.5=26.0.
claude-opus-4.5,apex_shortlist,47.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-opus-4.5=47.4.
gemini-3-pro,apex_shortlist,71.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gemini-3-pro=71.4.
doubao-seed-2.0-pro,apex_shortlist,82.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=82.1.
gpt-5.2,beyond_aime,86.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gpt-5.2=86.0.
claude-sonnet-4.5,beyond_aime,57.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-sonnet-4.5=57.0.
claude-opus-4.5,beyond_aime,69.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-opus-4.5=69.0.
gemini-3-pro,beyond_aime,83.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gemini-3-pro=83.0.
doubao-seed-2.0-pro,beyond_aime,86.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=86.5.
doubao-seed-2.0-pro,imo_answerbench,89.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=89.3.
gpt-5.2,aethercode,73.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gpt-5.2=73.8.
claude-sonnet-4.5,aethercode,16.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-sonnet-4.5=16.4.
claude-opus-4.5,aethercode,31.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-opus-4.5=31.6.
gemini-3-pro,aethercode,57.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gemini-3-pro=57.8.
doubao-seed-2.0-pro,aethercode,60.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=60.6.
gpt-5.2,livecodebench_v6,87.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gpt-5.2=87.7.
doubao-seed-2.0-pro,livecodebench_v6,87.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=87.8.
gpt-5.2,superchem,58.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gpt-5.2=58.0.
claude-sonnet-4.5,superchem,32.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-sonnet-4.5=32.4.
claude-opus-4.5,superchem,43.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-opus-4.5=43.2.
gemini-3-pro,superchem,63.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gemini-3-pro=63.2.
doubao-seed-2.0-pro,superchem,51.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=51.6.
gpt-5.2,babe,58.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gpt-5.2=58.1.
claude-sonnet-4.5,babe,44.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-sonnet-4.5=44.7.
claude-opus-4.5,babe,49.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-opus-4.5=49.3.
gemini-3-pro,babe,51.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gemini-3-pro=51.3.
doubao-seed-2.0-pro,babe,50.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=50.0.
gpt-5.2,phybench,74.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gpt-5.2=74.0.
claude-sonnet-4.5,phybench,48.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-sonnet-4.5=48.0.
claude-opus-4.5,phybench,69.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-opus-4.5=69.0.
gemini-3-pro,phybench,80.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gemini-3-pro=80.0.
doubao-seed-2.0-pro,phybench,74.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=74.0.
gpt-5.2,frontiersci_research,25.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gpt-5.2=25.0.
claude-sonnet-4.5,frontiersci_research,16.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-sonnet-4.5=16.7.
claude-opus-4.5,frontiersci_research,21.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-opus-4.5=21.7.
gemini-3-pro,frontiersci_research,15.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gemini-3-pro=15.0.
doubao-seed-2.0-pro,frontiersci_research,25.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=25.0.
gpt-5.2,frontiersci_olympiad,75.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gpt-5.2=75.0.
claude-sonnet-4.5,frontiersci_olympiad,60.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-sonnet-4.5=60.0.
claude-opus-4.5,frontiersci_olympiad,71.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-opus-4.5=71.0.
gemini-3-pro,frontiersci_olympiad,73.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gemini-3-pro=73.0.
doubao-seed-2.0-pro,frontiersci_olympiad,74.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=74.0.
doubao-seed-2.0-pro,arc_agi_1,85.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=85.4.
doubao-seed-2.0-pro,arc_agi_2,37.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=37.5.
gpt-5.2,korbench,79.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gpt-5.2=79.2.
claude-sonnet-4.5,korbench,73.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-sonnet-4.5=73.0.
claude-opus-4.5,korbench,77.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-opus-4.5=77.4.
gemini-3-pro,korbench,73.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gemini-3-pro=73.9.
doubao-seed-2.0-pro,korbench,77.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=77.5.
gpt-5.2,procbench,95.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gpt-5.2=95.0.
claude-sonnet-4.5,procbench,87.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-sonnet-4.5=87.5.
claude-opus-4.5,procbench,92.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-opus-4.5=92.5.
gemini-3-pro,procbench,90.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gemini-3-pro=90.0.
doubao-seed-2.0-pro,procbench,96.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=96.6.
gpt-5.2,mrcr_v2_8needle,89.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gpt-5.2=89.4.
claude-sonnet-4.5,mrcr_v2_8needle,47.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-sonnet-4.5=47.1.
claude-opus-4.5,mrcr_v2_8needle,56.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-opus-4.5=56.2.
gemini-3-pro,mrcr_v2_8needle,79.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gemini-3-pro=79.7.
doubao-seed-2.0-pro,mrcr_v2_8needle,54.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=54.0.
claude-sonnet-4.5,graphwalks_bfs_0k_128k,80.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-sonnet-4.5=80.5.
claude-opus-4.5,graphwalks_bfs_0k_128k,92.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-opus-4.5=92.0.
gemini-3-pro,graphwalks_bfs_0k_128k,79.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gemini-3-pro=79.9.
doubao-seed-2.0-pro,graphwalks_bfs_0k_128k,68.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=68.9.
claude-sonnet-4.5,graphwalks_parents_0k_128k,99.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-sonnet-4.5=99.0.
claude-opus-4.5,graphwalks_parents_0k_128k,96.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-opus-4.5=96.2.
gemini-3-pro,graphwalks_parents_0k_128k,99.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gemini-3-pro=99.7.
doubao-seed-2.0-pro,graphwalks_parents_0k_128k,97.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=97.6.
claude-sonnet-4.5,longbench_v2,62.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-sonnet-4.5=62.0.
doubao-seed-2.0-pro,longbench_v2,63.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=63.8.
gpt-5.2,frames,84.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gpt-5.2=84.0.
claude-sonnet-4.5,frames,78.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-sonnet-4.5=78.7.
claude-opus-4.5,frames,84.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-opus-4.5=84.7.
gemini-3-pro,frames,81.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gemini-3-pro=81.9.
doubao-seed-2.0-pro,frames,84.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=84.5.
gpt-5.2,der2_bench,69.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gpt-5.2=69.0.
claude-sonnet-4.5,der2_bench,58.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-sonnet-4.5=58.9.
claude-opus-4.5,der2_bench,60.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-opus-4.5=60.4.
gemini-3-pro,der2_bench,66.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gemini-3-pro=66.1.
doubao-seed-2.0-pro,der2_bench,58.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=58.2.
gpt-5.2,cl_bench,23.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gpt-5.2=23.9.
claude-sonnet-4.5,cl_bench,18.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-sonnet-4.5=18.1.
claude-opus-4.5,cl_bench,22.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-opus-4.5=22.6.
gemini-3-pro,cl_bench,15.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gemini-3-pro=15.6.
doubao-seed-2.0-pro,cl_bench,20.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=20.8.
doubao-seed-2.0-pro,global_piqa,92.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=92.3.
doubao-seed-2.0-pro,mmmlu,88.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=88.1.
gpt-5.2,disco_x,76.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gpt-5.2=76.3.
claude-sonnet-4.5,disco_x,70.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-sonnet-4.5=70.3.
claude-opus-4.5,disco_x,78.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-opus-4.5=78.6.
gemini-3-pro,disco_x,76.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gemini-3-pro=76.8.
doubao-seed-2.0-pro,disco_x,82.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=82.0.
claude-sonnet-4.5,multichallenge,57.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-sonnet-4.5=57.3.
doubao-seed-2.0-pro,multichallenge,68.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=68.3.
gpt-5.2,collie,96.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gpt-5.2=96.9.
claude-sonnet-4.5,collie,77.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-sonnet-4.5=77.3.
claude-opus-4.5,collie,79.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-opus-4.5=79.8.
gemini-3-pro,collie,95.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gemini-3-pro=95.0.
doubao-seed-2.0-pro,collie,93.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=93.9.
gpt-5.2,mars_bench,87.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gpt-5.2=87.9.
claude-sonnet-4.5,mars_bench,72.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-sonnet-4.5=72.9.
claude-opus-4.5,mars_bench,87.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-opus-4.5=87.7.
gemini-3-pro,mars_bench,85.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gemini-3-pro=85.6.
doubao-seed-2.0-pro,mars_bench,85.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=85.6.
gpt-5.2,inverse_ifeval,72.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gpt-5.2=72.3.
claude-sonnet-4.5,inverse_ifeval,69.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-sonnet-4.5=69.3.
claude-opus-4.5,inverse_ifeval,72.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-opus-4.5=72.4.
gemini-3-pro,inverse_ifeval,79.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gemini-3-pro=79.6.
doubao-seed-2.0-pro,inverse_ifeval,78.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=78.9.
gpt-5.2,longfact_objects,99.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gpt-5.2=99.2.
claude-sonnet-4.5,longfact_objects,98.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-sonnet-4.5=98.8.
claude-opus-4.5,longfact_objects,99.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-opus-4.5=99.0.
gemini-3-pro,longfact_objects,98.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gemini-3-pro=98.1.
doubao-seed-2.0-pro,longfact_objects,92.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=92.9.
gpt-5.2,longfact_concepts,99.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gpt-5.2=99.7.
claude-sonnet-4.5,longfact_concepts,98.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-sonnet-4.5=98.5.
claude-opus-4.5,longfact_concepts,98.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-opus-4.5=98.8.
gemini-3-pro,longfact_concepts,98.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gemini-3-pro=98.5.
doubao-seed-2.0-pro,longfact_concepts,92.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=92.8.
gpt-5.2,factscore,91.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gpt-5.2=91.9.
claude-sonnet-4.5,factscore,90.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-sonnet-4.5=90.6.
claude-opus-4.5,factscore,91.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: claude-opus-4.5=91.1.
gemini-3-pro,factscore,92.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: gemini-3-pro=92.6.
doubao-seed-2.0-pro,factscore,71.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card Table 3: doubao-seed-2.0-pro=71.2.
claude-opus-4.5,mathvista,80.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=80.6.
gpt-5.2,mathvista,83.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=83.1.
gemini-3-pro,mathvista,89.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=89.8.
doubao-seed-2.0-pro,mathvista,89.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=89.8.
claude-opus-4.5,dynamath,52.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=52.5.
gpt-5.2,dynamath,70.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=70.1.
gemini-3-pro,dynamath,63.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=63.3.
doubao-seed-2.0-pro,dynamath,68.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=68.9.
claude-opus-4.5,mathkangaroo,69.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=69.6.
gpt-5.2,mathkangaroo,86.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=86.9.
gemini-3-pro,mathkangaroo,84.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=84.4.
doubao-seed-2.0-pro,mathkangaroo,90.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=90.5.
claude-opus-4.5,mathcanvas,52.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=52.9.
gpt-5.2,mathcanvas,55.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=55.3.
gemini-3-pro,mathcanvas,58.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=58.8.
doubao-seed-2.0-pro,mathcanvas,61.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=61.9.
doubao-seed-2.0-pro,mmmu_pro,78.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=78.2.
claude-opus-4.5,emma,60.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=60.4.
gpt-5.2,emma,69.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=69.4.
gemini-3-pro,emma,66.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=66.5.
doubao-seed-2.0-pro,emma,72.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=72.0.
claude-opus-4.5,sfe,55.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=55.8.
gpt-5.2,sfe,50.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=50.1.
gemini-3-pro,sfe,61.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=61.9.
doubao-seed-2.0-pro,sfe,55.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=55.6.
claude-opus-4.5,hipho,81.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=81.8.
gpt-5.2,hipho,77.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=77.7.
gemini-3-pro,hipho,79.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=79.1.
doubao-seed-2.0-pro,hipho,74.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=74.1.
claude-opus-4.5,xlrs_macro,50.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=50.4.
gpt-5.2,xlrs_macro,49.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=49.9.
gemini-3-pro,xlrs_macro,51.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=51.7.
doubao-seed-2.0-pro,xlrs_macro,54.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=54.6.
claude-opus-4.5,phyx_openended,61.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=61.3.
gpt-5.2,phyx_openended,71.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=71.5.
gemini-3-pro,phyx_openended,71.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=71.0.
doubao-seed-2.0-pro,phyx_openended,72.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=72.1.
claude-opus-4.5,logicvista,68.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=68.9.
gpt-5.2,logicvista,81.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=81.0.
gemini-3-pro,logicvista,80.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=80.8.
doubao-seed-2.0-pro,logicvista,81.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=81.9.
claude-opus-4.5,vpct,29.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=29.0.
gpt-5.2,vpct,56.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=56.0.
gemini-3-pro,vpct,90.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=90.0.
doubao-seed-2.0-pro,vpct,76.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=76.0.
claude-opus-4.5,zerobench_main,4.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=4.0.
gpt-5.2,zerobench_main,11.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=11.0.
gemini-3-pro,zerobench_main,10.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=10.0.
doubao-seed-2.0-pro,zerobench_main,12.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=12.0.
claude-opus-4.5,zerobench_sub,30.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=30.8.
gpt-5.2,zerobench_sub,38.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=38.9.
gemini-3-pro,zerobench_sub,42.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=42.2.
doubao-seed-2.0-pro,zerobench_sub,47.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=47.6.
claude-opus-4.5,arcagi1_image,75.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=75.8.
gpt-5.2,arcagi1_image,93.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=93.1.
gemini-3-pro,arcagi1_image,69.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=69.4.
doubao-seed-2.0-pro,arcagi1_image,88.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=88.8.
claude-opus-4.5,arcagi2_image,26.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=26.1.
gpt-5.2,arcagi2_image,54.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=54.4.
gemini-3-pro,arcagi2_image,21.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=21.5.
doubao-seed-2.0-pro,arcagi2_image,43.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=43.3.
claude-opus-4.5,visulogic,27.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=27.6.
gpt-5.2,visulogic,37.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=37.0.
gemini-3-pro,visulogic,39.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=39.0.
doubao-seed-2.0-pro,visulogic,47.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=47.4.
claude-opus-4.5,vlms_are_biased,21.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=21.4.
gpt-5.2,vlms_are_biased,28.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=28.0.
gemini-3-pro,vlms_are_biased,50.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=50.6.
doubao-seed-2.0-pro,vlms_are_biased,77.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=77.4.
claude-opus-4.5,vlms_are_blind,77.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=77.2.
gpt-5.2,vlms_are_blind,84.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=84.2.
gemini-3-pro,vlms_are_blind,97.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=97.5.
doubao-seed-2.0-pro,vlms_are_blind,98.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=98.6.
claude-opus-4.5,visfactor,24.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=24.5.
gpt-5.2,visfactor,33.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=33.6.
gemini-3-pro,visfactor,45.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=45.8.
doubao-seed-2.0-pro,visfactor,36.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=36.8.
claude-opus-4.5,realworldqa,75.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=75.9.
gpt-5.2,realworldqa,82.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=82.1.
gemini-3-pro,realworldqa,84.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=84.7.
doubao-seed-2.0-pro,realworldqa,86.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=86.0.
claude-opus-4.5,babyvision,16.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=16.2.
gpt-5.2,babyvision,37.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=37.4.
gemini-3-pro,babyvision,49.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=49.7.
doubao-seed-2.0-pro,babyvision,60.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=60.6.
doubao-seed-2.0-pro,simplevqa,71.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=71.4.
claude-opus-4.5,hallusionbench,65.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=65.3.
gpt-5.2,hallusionbench,67.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=67.7.
gemini-3-pro,hallusionbench,69.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=69.9.
doubao-seed-2.0-pro,hallusionbench,68.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=68.0.
claude-opus-4.5,mme_cc,25.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=25.2.
gpt-5.2,mme_cc,44.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=44.4.
gemini-3-pro,mme_cc,56.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=56.9.
doubao-seed-2.0-pro,mme_cc,57.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=57.0.
claude-opus-4.5,mmstar,73.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=73.9.
gpt-5.2,mmstar,78.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=78.2.
gemini-3-pro,mmstar,83.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=83.1.
doubao-seed-2.0-pro,mmstar,83.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=83.0.
claude-opus-4.5,muirbench,78.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=78.9.
gpt-5.2,muirbench,77.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=77.4.
gemini-3-pro,muirbench,78.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=78.2.
doubao-seed-2.0-pro,muirbench,81.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=81.8.
claude-opus-4.5,mtvqa,53.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=53.1.
gpt-5.2,mtvqa,48.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=48.5.
gemini-3-pro,mtvqa,50.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=50.8.
doubao-seed-2.0-pro,mtvqa,51.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=51.1.
doubao-seed-2.0-pro,worldvqa,49.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=49.9.
claude-opus-4.5,vibe_eval,70.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=70.3.
gpt-5.2,vibe_eval,73.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=73.1.
gemini-3-pro,vibe_eval,77.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=77.7.
doubao-seed-2.0-pro,vibe_eval,81.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=81.4.
claude-opus-4.5,viverbench,72.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=72.4.
gpt-5.2,viverbench,74.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=74.8.
gemini-3-pro,viverbench,75.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=75.9.
doubao-seed-2.0-pro,viverbench,75.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=75.9.
claude-opus-4.5,countbench,90.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=90.3.
gpt-5.2,countbench,91.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=91.2.
gemini-3-pro,countbench,97.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=97.3.
doubao-seed-2.0-pro,countbench,95.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=95.5.
claude-opus-4.5,fsc_147,20.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=20.9.
gpt-5.2,fsc_147,21.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=21.1.
gemini-3-pro,fsc_147,12.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=12.1.
doubao-seed-2.0-pro,fsc_147,11.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=11.3.
gemini-3-pro,point_bench,85.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=85.5.
doubao-seed-2.0-pro,point_bench,81.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=81.4.
claude-opus-4.5,blink,68.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=68.1.
gpt-5.2,blink,70.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=70.3.
gemini-3-pro,blink,77.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=77.1.
doubao-seed-2.0-pro,blink,79.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=79.5.
claude-opus-4.5,mmsibench_circular,20.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=20.2.
gpt-5.2,mmsibench_circular,26.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=26.1.
gemini-3-pro,mmsibench_circular,25.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=25.4.
doubao-seed-2.0-pro,mmsibench_circular,32.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=32.5.
claude-opus-4.5,treebench,53.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=53.6.
gpt-5.2,treebench,58.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=58.8.
gemini-3-pro,treebench,62.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=62.7.
doubao-seed-2.0-pro,treebench,64.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=64.7.
gpt-5.2,refspatialbench,25.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=25.5.
gemini-3-pro,refspatialbench,65.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=65.5.
doubao-seed-2.0-pro,refspatialbench,72.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=72.6.
claude-opus-4.5,da_2k,70.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=70.3.
gpt-5.2,da_2k,78.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=78.9.
gemini-3-pro,da_2k,82.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=82.1.
doubao-seed-2.0-pro,da_2k,92.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=92.3.
claude-opus-4.5,all_angles,63.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=63.1.
gpt-5.2,all_angles,71.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=71.5.
gemini-3-pro,all_angles,73.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=73.5.
doubao-seed-2.0-pro,all_angles,72.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=72.1.
claude-opus-4.5,erqa,48.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=48.3.
gpt-5.2,erqa,59.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=59.8.
gemini-3-pro,erqa,70.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=70.5.
doubao-seed-2.0-pro,erqa,68.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=68.5.
gpt-5.2,chartqapro,67.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=67.6.
gemini-3-pro,chartqapro,69.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=69.0.
doubao-seed-2.0-pro,chartqapro,71.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=71.2.
claude-opus-4.5,ocrbench_v2,55.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=55.5.
gpt-5.2,ocrbench_v2,55.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=55.6.
gemini-3-pro,ocrbench_v2,63.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=63.3.
doubao-seed-2.0-pro,ocrbench_v2,62.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=62.5.
doubao-seed-2.0-pro,omnidocbench_1.5,0.099,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=0.099.
claude-opus-4.5,charxiv_descriptive,92.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=92.7.
gpt-5.2,charxiv_descriptive,93.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=93.8.
gemini-3-pro,charxiv_descriptive,94.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=94.4.
doubao-seed-2.0-pro,charxiv_descriptive,93.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=93.5.
doubao-seed-2.0-pro,charxiv_reasoning,80.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=80.5.
claude-opus-4.5,dude,55.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=55.6.
gpt-5.2,dude,68.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=68.2.
gemini-3-pro,dude,70.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=70.1.
doubao-seed-2.0-pro,dude,72.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=72.4.
gemini-3-pro,mmlongbench,73.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=73.6.
doubao-seed-2.0-pro,mmlongbench,74.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=74.8.
gemini-3-pro,longdocurl,72.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=72.0.
doubao-seed-2.0-pro,longdocurl,74.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=74.7.
gemini-3-pro,mmlongbench_doc,59.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=59.5.
doubao-seed-2.0-pro,mmlongbench_doc,61.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=61.4.
doubao-seed-2.0-pro,mmvu,78.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=78.2.
gemini-3-pro,videosimpleqa,71.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=71.9.
doubao-seed-2.0-pro,videosimpleqa,71.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=71.9.
gemini-3-pro,videoreasonbench,59.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=59.5.
doubao-seed-2.0-pro,videoreasonbench,77.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=77.8.
gemini-3-pro,morse_500,33.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=33.0.
doubao-seed-2.0-pro,morse_500,37.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=37.4.
gemini-3-pro,videoholmes,64.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=64.2.
doubao-seed-2.0-pro,videoholmes,67.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=67.4.
gemini-3-pro,minerva,65.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=65.0.
doubao-seed-2.0-pro,minerva,66.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=66.5.
gemini-3-pro,tvbench,71.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=71.1.
doubao-seed-2.0-pro,tvbench,75.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=75.0.
gemini-3-pro,contphy,58.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=58.0.
doubao-seed-2.0-pro,contphy,67.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=67.4.
gemini-3-pro,tempcompass,88.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=88.0.
doubao-seed-2.0-pro,tempcompass,89.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=89.6.
gemini-3-pro,egotempo,65.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=65.4.
doubao-seed-2.0-pro,egotempo,71.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=71.8.
doubao-seed-2.0-pro,motionbench,75.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=75.2.
gemini-3-pro,tomato,59.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=59.6.
doubao-seed-2.0-pro,tomato,59.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=59.9.
doubao-seed-2.0-pro,video_mme,89.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=89.5.
gemini-3-pro,cgbench,65.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=65.5.
doubao-seed-2.0-pro,cgbench,65.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=65.0.
doubao-seed-2.0-pro,longvideobench,80.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=80.3.
gemini-3-pro,videoeval_pro,52.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=52.7.
doubao-seed-2.0-pro,videoeval_pro,48.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=48.0.
doubao-seed-2.0-pro,lvbench,76.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=76.4.
gemini-3-pro,crossvid,53.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=53.0.
doubao-seed-2.0-pro,crossvid,60.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=60.3.
gemini-3-pro,ovbench,62.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=62.7.
doubao-seed-2.0-pro,ovbench,69.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=69.2.
gemini-3-pro,livesports_3k,74.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=74.5.
doubao-seed-2.0-pro,livesports_3k,78.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=78.0.
gemini-3-pro,ovobench,70.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=70.1.
doubao-seed-2.0-pro,ovobench,77.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=77.0.
gemini-3-pro,odvbench,63.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=63.6.
doubao-seed-2.0-pro,odvbench,72.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=72.5.
gemini-3-pro,vispeak,89.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=89.0.
doubao-seed-2.0-pro,vispeak,78.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=78.5.
mistral-small-3.1,math,69.3,https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Instruct-2503,"{""harness"":""mistral-eval"",""judge"":""rule-based"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.0""}",True,model_card,verified,"HF MC Table 1: MATH = 69.30% (4-shot CoT, official Mistral eval)."
mistral-small-3.1,mmmu,64.0,https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Instruct-2503,"{""harness"":""mistral-eval"",""judge"":""rule-based"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.0""}",True,model_card,verified,HF MC Table 2: MMMU = 64.00%.
mistral-small-3.1,mmmu_pro,49.25,https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Instruct-2503,"{""harness"":""mistral-eval"",""judge"":""rule-based"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.0""}",True,model_card,verified,HF MC Table 2: MMMU PRO = 49.25%.
mistral-small-3.1,mathvista,68.91,https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Instruct-2503,"{""harness"":""mistral-eval"",""judge"":""rule-based"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.0""}",True,model_card,verified,HF MC Table 2: Mathvista = 68.91%.
mistral-small-3.1,longbench_v2,37.18,https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Instruct-2503,"{""harness"":""mistral-eval"",""judge"":""rule-based"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.0""}",True,model_card,verified,HF MC Table 4: LongBench v2 = 37.18%.
gpt-4o-mini,math_500,73.0,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,Phi-4 paper Table 1 (gpt-4o-mini column): math_500=73.0
gpt-4o-mini,humaneval,86.2,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,Phi-4 paper Table 1 (gpt-4o-mini column): humaneval=86.2
phi-4,mgsm,80.6,https://arxiv.org/abs/2412.08905,"{""effort"":""n/a"",""harness"":""simple-evals"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.5"",""tools"":""none""}",True,tech_report,verified,Phi-4 paper Table 1 (phi-4 column): mgsm=80.6
gpt-4o-mini,mgsm,86.5,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,Phi-4 paper Table 1 (gpt-4o-mini column): mgsm=86.5
qwen2.5-72b-instruct,mgsm,87.3,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,Phi-4 paper Table 1 (qwen2.5-72b-instruct column): mgsm=87.3
gpt-4o-0513,mgsm,90.4,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,Phi-4 paper Table 1 (gpt-4o-0513 column): mgsm=90.4
gpt-4o-mini,simpleqa,9.9,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,Phi-4 paper Table 1 (gpt-4o-mini column): simpleqa=9.9
phi-4,drop,75.5,https://arxiv.org/abs/2412.08905,"{""effort"":""n/a"",""harness"":""simple-evals"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.5"",""tools"":""none""}",True,tech_report,verified,Phi-4 paper Table 1 (phi-4 column): drop=75.5
gpt-4o-mini,drop,79.3,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,Phi-4 paper Table 1 (gpt-4o-mini column): drop=79.3
gpt-4o-mini,mmlu_pro,63.4,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,Phi-4 paper Table 1 (gpt-4o-mini column): mmlu_pro=63.4
phi-4,humaneval_plus,82.8,https://arxiv.org/abs/2412.08905,"{""effort"":""n/a"",""harness"":""simple-evals"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.5"",""tools"":""none""}",True,tech_report,verified,Phi-4 paper Table 1 (phi-4 column): humaneval_plus=82.8
gpt-4o-mini,humaneval_plus,82.0,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,Phi-4 paper Table 1 (gpt-4o-mini column): humaneval_plus=82.0
qwen2.5-72b-instruct,humaneval_plus,78.4,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,Phi-4 paper Table 1 (qwen2.5-72b-instruct column): humaneval_plus=78.4
gpt-4o-0513,humaneval_plus,88.0,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,Phi-4 paper Table 1 (gpt-4o-0513 column): humaneval_plus=88.0
gpt-4o-mini,arena_hard,76.2,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,Phi-4 paper Table 1 (gpt-4o-mini column): arena_hard=76.2
qwen2.5-72b-instruct,arena_hard,78.4,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,Phi-4 paper Table 1 (qwen2.5-72b-instruct column): arena_hard=78.4
gpt-4o-0513,arena_hard,75.6,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,Phi-4 paper Table 1 (gpt-4o-0513 column): arena_hard=75.6
phi-4,livebench,47.6,https://arxiv.org/abs/2412.08905,"{""effort"":""n/a"",""harness"":""simple-evals"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.5"",""tools"":""none""}",True,tech_report,verified,Phi-4 paper Table 1 (phi-4 column): livebench=47.6
gpt-4o-mini,livebench,48.1,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,Phi-4 paper Table 1 (gpt-4o-mini column): livebench=48.1
qwen2.5-72b-instruct,livebench,55.3,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,Phi-4 paper Table 1 (qwen2.5-72b-instruct column): livebench=55.3
gpt-4o-0513,livebench,57.6,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,Phi-4 paper Table 1 (gpt-4o-0513 column): livebench=57.6
o1-high,aime_2025,71.4,https://arxiv.org/abs/2504.21318,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,Phi-4-reasoning paper Table 1 (o1-high column): aime_2025=71.4
o1-high,hmmt_feb_2025,48.3,https://arxiv.org/abs/2504.21318,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,Phi-4-reasoning paper Table 1 (o1-high column): hmmt_2025=48.3
claude-3.7-sonnet,hmmt_feb_2025,31.7,https://arxiv.org/abs/2504.21318,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,Phi-4-reasoning paper Table 1 (claude-3.7-sonnet column): hmmt_2025=31.7
phi-4-reasoning,omnimath,76.6,https://arxiv.org/abs/2504.21318,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.8"",""tools"":""none""}",True,tech_report,verified,Phi-4-reasoning paper Table 1 (phi-4-reasoning column): omnimath=76.6
phi-4-reasoning-plus,omnimath,81.9,https://arxiv.org/abs/2504.21318,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.8"",""tools"":""none""}",True,tech_report,verified,Phi-4-reasoning paper Table 1 (phi-4-reasoning-plus column): omnimath=81.9
deepseek-r1-distill-llama-70b,omnimath,63.4,https://arxiv.org/abs/2504.21318,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,Phi-4-reasoning paper Table 1 (deepseek-r1-distill-llama-70b column): omnimath=63.4
deepseek-r1,omnimath,85.0,https://arxiv.org/abs/2504.21318,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,Phi-4-reasoning paper Table 1 (deepseek-r1 column): omnimath=85.0
o1-high,omnimath,67.5,https://arxiv.org/abs/2504.21318,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,Phi-4-reasoning paper Table 1 (o1-high column): omnimath=67.5
o3-mini-high,omnimath,74.6,https://arxiv.org/abs/2504.21318,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,Phi-4-reasoning paper Table 1 (o3-mini-high column): omnimath=74.6
claude-3.7-sonnet,omnimath,54.6,https://arxiv.org/abs/2504.21318,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,Phi-4-reasoning paper Table 1 (claude-3.7-sonnet column): omnimath=54.6
o1-high,livecodebench,63.4,https://arxiv.org/abs/2504.21318,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,Phi-4-reasoning paper Table 1 (o1-high column): livecodebench=63.4
o1-high,codeforces_rating,1891.0,https://arxiv.org/abs/2504.21318,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,Phi-4-reasoning paper Table 1 (o1-high column): codeforces_rating=1891
phi-4,flenqa_3k,82.0,https://arxiv.org/abs/2504.21318,"{""effort"":""n/a"",""harness"":""simple-evals"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.5"",""tools"":""none""}",True,tech_report,verified,Phi-4-reasoning paper Table 2 (phi-4 column): flenqa_3k=82.0
phi-4-reasoning,flenqa_3k,97.7,https://arxiv.org/abs/2504.21318,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.8"",""tools"":""none""}",True,tech_report,verified,Phi-4-reasoning paper Table 2 (phi-4-reasoning column): flenqa_3k=97.7
phi-4-reasoning-plus,flenqa_3k,97.9,https://arxiv.org/abs/2504.21318,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.8"",""tools"":""none""}",True,tech_report,verified,Phi-4-reasoning paper Table 2 (phi-4-reasoning-plus column): flenqa_3k=97.9
o3-mini-high,flenqa_3k,96.8,https://arxiv.org/abs/2504.21318,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,Phi-4-reasoning paper Table 2 (o3-mini-high column): flenqa_3k=96.8
gpt-4o-0513,flenqa_3k,90.8,https://arxiv.org/abs/2504.21318,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,Phi-4-reasoning paper Table 2 (gpt-4o-0513 column): flenqa_3k=90.8
phi-4-reasoning,humaneval_plus,92.9,https://arxiv.org/abs/2504.21318,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.8"",""tools"":""none""}",True,tech_report,verified,Phi-4-reasoning paper Table 2 (phi-4-reasoning column): humaneval_plus=92.9
phi-4-reasoning-plus,humaneval_plus,92.3,https://arxiv.org/abs/2504.21318,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.8"",""tools"":""none""}",True,tech_report,verified,Phi-4-reasoning paper Table 2 (phi-4-reasoning-plus column): humaneval_plus=92.3
o3-mini-high,humaneval_plus,94.0,https://arxiv.org/abs/2504.21318,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,Phi-4-reasoning paper Table 2 (o3-mini-high column): humaneval_plus=94.0
phi-4,phibench,58.2,https://arxiv.org/abs/2504.21318,"{""effort"":""n/a"",""harness"":""simple-evals"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.5"",""tools"":""none""}",True,tech_report,verified,Phi-4-reasoning paper Table 2 (phi-4 column): phibench=58.2
phi-4-reasoning,phibench,70.6,https://arxiv.org/abs/2504.21318,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.8"",""tools"":""none""}",True,tech_report,verified,Phi-4-reasoning paper Table 2 (phi-4-reasoning column): phibench=70.6
phi-4-reasoning-plus,phibench,74.2,https://arxiv.org/abs/2504.21318,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.8"",""tools"":""none""}",True,tech_report,verified,Phi-4-reasoning paper Table 2 (phi-4-reasoning-plus column): phibench=74.2
o3-mini-high,phibench,78.0,https://arxiv.org/abs/2504.21318,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,Phi-4-reasoning paper Table 2 (o3-mini-high column): phibench=78.0
gpt-4o-0513,phibench,73.1,https://arxiv.org/abs/2504.21318,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,Phi-4-reasoning paper Table 2 (gpt-4o-0513 column): phibench=73.1
phi-4-mini,bigbench_hard,70.4,https://huggingface.co/microsoft/Phi-4-mini-instruct,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,phi-4-mini HF model card: bigbench_hard=70.4
phi-4-mini,arc_challenge,83.7,https://huggingface.co/microsoft/Phi-4-mini-instruct,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,phi-4-mini HF model card: arc_challenge=83.7
phi-4-mini,gpqa_main,25.2,https://huggingface.co/microsoft/Phi-4-mini-instruct,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,phi-4-mini HF model card: gpqa=25.2
phi-4-mini,math,64.0,https://huggingface.co/microsoft/Phi-4-mini-instruct,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,phi-4-mini HF model card: math=64.0
phi-4-mini,multilingual_mmlu,49.3,https://huggingface.co/microsoft/Phi-4-mini-instruct,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,phi-4-mini HF model card: multilingual_mmlu=49.3
phi-4-mini,mgsm,63.9,https://huggingface.co/microsoft/Phi-4-mini-instruct,"{""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,model_card,verified,phi-4-mini HF model card: mgsm=63.9
claude-sonnet-4.5,swelancer,45.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-sonnet-4.5=45.7.
claude-opus-4.5,swelancer,56.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=56.1.
gemini-3-pro,swelancer,44.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=44.3.
doubao-seed-2.0-pro,swelancer,49.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=49.4.
gpt-5.2,multi_swe_bench,47.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=47.7.
claude-opus-4.5,multi_swe_bench,52.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=52.8.
gemini-3-pro,multi_swe_bench,50.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=50.2.
doubao-seed-2.0-pro,multi_swe_bench,45.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=45.2.
doubao-seed-2.0-pro,swe_bench_pro,46.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=46.9.
doubao-seed-2.0-pro,swe_bench_multilingual,71.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=71.7.
doubao-seed-2.0-pro,scicode,48.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=48.5.
gpt-5.2,swe_evo,12.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=12.5.
claude-sonnet-4.5,swe_evo,16.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-sonnet-4.5=16.7.
claude-opus-4.5,swe_evo,27.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=27.1.
gemini-3-pro,swe_evo,8.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=8.9.
doubao-seed-2.0-pro,swe_evo,8.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=8.5.
gpt-5.2,aider_polyglot_diff,91.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=91.1.
claude-sonnet-4.5,aider_polyglot_diff,82.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-sonnet-4.5=82.2.
claude-opus-4.5,aider_polyglot_diff,92.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=92.4.
gemini-3-pro,aider_polyglot_diff,94.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=94.2.
doubao-seed-2.0-pro,aider_polyglot_diff,80.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=80.0.
gpt-5.2,artifactsbench,71.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=71.1.
claude-opus-4.5,artifactsbench,68.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=68.5.
gemini-3-pro,artifactsbench,58.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=58.4.
doubao-seed-2.0-pro,artifactsbench,66.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=66.6.
gpt-5.2,codesimpleqa,62.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=62.3.
claude-sonnet-4.5,codesimpleqa,59.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-sonnet-4.5=59.6.
claude-opus-4.5,codesimpleqa,63.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=63.0.
gemini-3-pro,codesimpleqa,54.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=54.7.
doubao-seed-2.0-pro,codesimpleqa,58.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=58.0.
gpt-5.2,spreadsheetbench_verified,69.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=69.9.
claude-sonnet-4.5,spreadsheetbench_verified,75.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-sonnet-4.5=75.9.
claude-opus-4.5,spreadsheetbench_verified,78.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=78.6.
gemini-3-pro,spreadsheetbench_verified,70.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=70.8.
doubao-seed-2.0-pro,spreadsheetbench_verified,79.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=79.1.
doubao-seed-2.0-pro,browsecomp_zh,82.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=82.4.
gpt-5.2,hle_text,45.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=45.5.
claude-sonnet-4.5,hle_text,32.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-sonnet-4.5=32.0.
claude-opus-4.5,hle_text,43.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=43.2.
gemini-3-pro,hle_text,46.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=46.9.
doubao-seed-2.0-pro,hle_text,54.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=54.2.
gpt-5.2,hle_verified,68.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=68.5.
claude-sonnet-4.5,hle_verified,37.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-sonnet-4.5=37.6.
claude-opus-4.5,hle_verified,56.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=56.6.
gemini-3-pro,hle_verified,67.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=67.5.
doubao-seed-2.0-pro,hle_verified,73.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=73.6.
gpt-5.2,widesearch,76.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=76.8.
claude-sonnet-4.5,widesearch,65.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-sonnet-4.5=65.1.
doubao-seed-2.0-pro,widesearch,74.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=74.7.
gpt-5.2,finsearchcomp,73.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=73.8.
claude-sonnet-4.5,finsearchcomp,58.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-sonnet-4.5=58.6.
claude-opus-4.5,finsearchcomp,66.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=66.2.
gemini-3-pro,finsearchcomp,52.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=52.7.
doubao-seed-2.0-pro,finsearchcomp,70.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=70.2.
claude-sonnet-4.5,deepsearchqa_acc,36.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-sonnet-4.5=36.3.
doubao-seed-2.0-pro,deepsearchqa_acc,77.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=77.4.
claude-sonnet-4.5,seal_0,53.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-sonnet-4.5=53.4.
doubao-seed-2.0-pro,seal_0,49.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=49.5.
doubao-seed-2.0-pro,tau2_bench_retail,90.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=90.4.
doubao-seed-2.0-pro,tau2_bench_telecom,94.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=94.2.
claude-sonnet-4.5,mcpmark,32.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-sonnet-4.5=32.1.
doubao-seed-2.0-pro,mcpmark,54.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=54.7.
gpt-5.2,bfcl_v4,65.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=65.9.
claude-sonnet-4.5,bfcl_v4,72.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-sonnet-4.5=72.9.
claude-opus-4.5,bfcl_v4,76.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=76.5.
gemini-3-pro,bfcl_v4,71.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=71.0.
doubao-seed-2.0-pro,bfcl_v4,73.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=73.4.
gpt-5.2,vitabench,41.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=41.8.
claude-sonnet-4.5,vitabench,40.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-sonnet-4.5=40.8.
claude-opus-4.5,vitabench,55.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=55.3.
gemini-3-pro,vitabench,48.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=48.8.
doubao-seed-2.0-pro,vitabench,47.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=47.0.
gpt-5.2,deepconsult,54.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=54.3.
claude-sonnet-4.5,deepconsult,55.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-sonnet-4.5=55.8.
claude-opus-4.5,deepconsult,61.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=61.0.
gemini-3-pro,deepconsult,48.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=48.0.
doubao-seed-2.0-pro,deepconsult,61.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=61.1.
gpt-5.2,deepresearchbench,52.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=52.2.
claude-sonnet-4.5,deepresearchbench,47.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-sonnet-4.5=47.2.
claude-opus-4.5,deepresearchbench,50.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=50.6.
gemini-3-pro,deepresearchbench,49.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=49.6.
doubao-seed-2.0-pro,deepresearchbench,53.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=53.3.
gpt-5.2,researchrubrics,42.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=42.3.
claude-sonnet-4.5,researchrubrics,38.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-sonnet-4.5=38.6.
claude-opus-4.5,researchrubrics,45.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=45.0.
gemini-3-pro,researchrubrics,37.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=37.7.
doubao-seed-2.0-pro,researchrubrics,50.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=50.7.
gpt-5.2,minedojo_verified,18.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=18.3.
gemini-3-pro,minedojo_verified,23.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=23.3.
doubao-seed-2.0-pro,minedojo_verified,49.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=49.0.
gpt-5.2,mm_browsecomp,26.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=26.3.
gemini-3-pro,mm_browsecomp,25.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=25.0.
doubao-seed-2.0-pro,mm_browsecomp,48.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=48.8.
gpt-5.2,hle_vl,31.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=31.0.
gemini-3-pro,hle_vl,36.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=36.0.
doubao-seed-2.0-pro,hle_vl,39.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=39.2.
gpt-5.2,biobench,58.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=58.1.
claude-sonnet-4.5,biobench,44.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-sonnet-4.5=44.7.
claude-opus-4.5,biobench,49.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=49.3.
gemini-3-pro,biobench,51.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=51.3.
doubao-seed-2.0-pro,biobench,53.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=53.5.
gpt-5.2,ainstein_bench,41.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=41.3.
claude-sonnet-4.5,ainstein_bench,33.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-sonnet-4.5=33.7.
claude-opus-4.5,ainstein_bench,44.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=44.0.
gemini-3-pro,ainstein_bench,42.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=42.8.
doubao-seed-2.0-pro,ainstein_bench,47.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=47.7.
gpt-5.2,nl2repo_bench,49.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=49.3.
claude-sonnet-4.5,nl2repo_bench,39.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-sonnet-4.5=39.9.
claude-opus-4.5,nl2repo_bench,43.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=43.2.
gemini-3-pro,nl2repo_bench,34.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=34.2.
doubao-seed-2.0-pro,nl2repo_bench,27.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=27.9.
gpt-5.2,nl2repo_pass1,8.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=8.0.
claude-sonnet-4.5,nl2repo_pass1,3.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-sonnet-4.5=3.0.
claude-opus-4.5,nl2repo_pass1,3.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=3.0.
gemini-3-pro,nl2repo_pass1,4.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=4.0.
doubao-seed-2.0-pro,nl2repo_pass1,3.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=3.0.
gpt-5.2,tob_complex_workflows,45.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=45.0.
claude-sonnet-4.5,tob_complex_workflows,61.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-sonnet-4.5=61.0.
claude-opus-4.5,tob_complex_workflows,64.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=64.8.
gemini-3-pro,tob_complex_workflows,69.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=69.2.
doubao-seed-2.0-pro,tob_complex_workflows,64.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=64.7.
gpt-5.2,tob_referenceqa,63.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=63.6.
claude-sonnet-4.5,tob_referenceqa,58.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-sonnet-4.5=58.9.
claude-opus-4.5,tob_referenceqa,67.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=67.9.
gemini-3-pro,tob_referenceqa,68.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=68.3.
doubao-seed-2.0-pro,tob_referenceqa,72.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=72.4.
gpt-5.2,gdpval_diamond,26.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=26.9.
claude-sonnet-4.5,gdpval_diamond,15.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-sonnet-4.5=15.2.
claude-opus-4.5,gdpval_diamond,20.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=20.7.
gemini-3-pro,gdpval_diamond,19.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=19.4.
doubao-seed-2.0-pro,gdpval_diamond,21.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=21.3.
gpt-5.2,xpert_bench,53.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=53.3.
claude-sonnet-4.5,xpert_bench,44.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-sonnet-4.5=44.7.
claude-opus-4.5,xpert_bench,50.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=50.5.
gemini-3-pro,xpert_bench,53.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=53.1.
doubao-seed-2.0-pro,xpert_bench,64.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=64.5.
gpt-5.2,tob_k12_education,61.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=61.6.
claude-sonnet-4.5,tob_k12_education,50.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-sonnet-4.5=50.1.
claude-opus-4.5,tob_k12_education,56.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=56.2.
gemini-3-pro,tob_k12_education,59.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=59.4.
doubao-seed-2.0-pro,tob_k12_education,62.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=62.8.
gpt-5.2,tob_compositional_tasks,51.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=51.5.
claude-sonnet-4.5,tob_compositional_tasks,57.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-sonnet-4.5=57.3.
claude-opus-4.5,tob_compositional_tasks,63.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=63.6.
gemini-3-pro,tob_compositional_tasks,64.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=64.8.
doubao-seed-2.0-pro,tob_compositional_tasks,59.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=59.1.
gpt-5.2,tob_text_classification,62.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=62.1.
claude-sonnet-4.5,tob_text_classification,64.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-sonnet-4.5=64.5.
claude-opus-4.5,tob_text_classification,63.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=63.9.
gemini-3-pro,tob_text_classification,67.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=67.5.
doubao-seed-2.0-pro,tob_text_classification,69.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=69.0.
gpt-5.2,tob_information_extraction,44.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=44.7.
claude-sonnet-4.5,tob_information_extraction,48.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-sonnet-4.5=48.3.
claude-opus-4.5,tob_information_extraction,50.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=50.1.
gemini-3-pro,tob_information_extraction,49.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=49.0.
doubao-seed-2.0-pro,tob_information_extraction,52.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=52.0.
gpt-5.2,world_travel_vlm,19.33,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=19.33.
claude-sonnet-4.5,world_travel_vlm,2.67,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-sonnet-4.5=2.67.
claude-opus-4.5,world_travel_vlm,14.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=14.0.
gemini-3-pro,world_travel_vlm,8.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=8.0.
doubao-seed-2.0-pro,world_travel_vlm,12.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=12.0.
gpt-5.2,world_travel_text,32.67,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gpt-5.2=32.67.
claude-sonnet-4.5,world_travel_text,10.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-sonnet-4.5=10.0.
claude-opus-4.5,world_travel_text,21.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: claude-opus-4.5=21.3.
gemini-3-pro,world_travel_text,14.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: gemini-3-pro=14.7.
doubao-seed-2.0-pro,world_travel_text,23.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""varies""}",True,model_card,verified,Doubao Seed 2.0 model card: doubao-seed-2.0-pro=23.3.
phi-3-14b,mmlu,77.9,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4 paper Table 1 baseline (phi-3-14b column): mmlu=77.9. Third-party Microsoft self-test, mc=false."
qwen-2.5-14b-instruct,mmlu,79.9,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4 paper Table 1 baseline (qwen-2.5-14b-instruct column): mmlu=79.9. Third-party Microsoft self-test, mc=false."
llama-3.3-70b-instruct,mmlu,86.3,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4 paper Table 1 baseline (llama-3.3-70b-instruct column): mmlu=86.3. Third-party Microsoft self-test, mc=false."
phi-3-14b,gpqa_diamond,31.2,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4 paper Table 1 baseline (phi-3-14b column): gpqa_diamond=31.2. Third-party Microsoft self-test, mc=false."
qwen-2.5-14b-instruct,gpqa_diamond,42.9,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4 paper Table 1 baseline (qwen-2.5-14b-instruct column): gpqa_diamond=42.9. Third-party Microsoft self-test, mc=false."
llama-3.3-70b-instruct,gpqa_diamond,49.1,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4 paper Table 1 baseline (llama-3.3-70b-instruct column): gpqa_diamond=49.1. Third-party Microsoft self-test, mc=false."
phi-3-14b,math_500,44.6,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4 paper Table 1 baseline (phi-3-14b column): math_500=44.6. Third-party Microsoft self-test, mc=false."
qwen-2.5-14b-instruct,math_500,75.6,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4 paper Table 1 baseline (qwen-2.5-14b-instruct column): math_500=75.6. Third-party Microsoft self-test, mc=false."
llama-3.3-70b-instruct,math_500,66.31,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4 paper Table 1 baseline (llama-3.3-70b-instruct column): math_500=66.31. Third-party Microsoft self-test, mc=false."
phi-3-14b,humaneval,67.8,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4 paper Table 1 baseline (phi-3-14b column): humaneval=67.8. Third-party Microsoft self-test, mc=false."
qwen-2.5-14b-instruct,humaneval,72.1,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4 paper Table 1 baseline (qwen-2.5-14b-instruct column): humaneval=72.1. Third-party Microsoft self-test, mc=false."
llama-3.3-70b-instruct,humaneval,78.91,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4 paper Table 1 baseline (llama-3.3-70b-instruct column): humaneval=78.91. Third-party Microsoft self-test, mc=false."
phi-3-14b,mgsm,53.5,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4 paper Table 1 baseline (phi-3-14b column): mgsm=53.5. Third-party Microsoft self-test, mc=false."
qwen-2.5-14b-instruct,mgsm,79.6,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4 paper Table 1 baseline (qwen-2.5-14b-instruct column): mgsm=79.6. Third-party Microsoft self-test, mc=false."
llama-3.3-70b-instruct,mgsm,89.1,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4 paper Table 1 baseline (llama-3.3-70b-instruct column): mgsm=89.1. Third-party Microsoft self-test, mc=false."
phi-3-14b,simpleqa,7.6,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4 paper Table 1 baseline (phi-3-14b column): simpleqa=7.6. Third-party Microsoft self-test, mc=false."
qwen-2.5-14b-instruct,simpleqa,5.4,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4 paper Table 1 baseline (qwen-2.5-14b-instruct column): simpleqa=5.4. Third-party Microsoft self-test, mc=false."
llama-3.3-70b-instruct,simpleqa,20.9,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4 paper Table 1 baseline (llama-3.3-70b-instruct column): simpleqa=20.9. Third-party Microsoft self-test, mc=false."
phi-3-14b,drop,68.3,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4 paper Table 1 baseline (phi-3-14b column): drop=68.3. Third-party Microsoft self-test, mc=false."
qwen-2.5-14b-instruct,drop,85.5,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4 paper Table 1 baseline (qwen-2.5-14b-instruct column): drop=85.5. Third-party Microsoft self-test, mc=false."
llama-3.3-70b-instruct,drop,90.2,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4 paper Table 1 baseline (llama-3.3-70b-instruct column): drop=90.2. Third-party Microsoft self-test, mc=false."
phi-3-14b,mmlu_pro,51.3,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4 paper Table 1 baseline (phi-3-14b column): mmlu_pro=51.3. Third-party Microsoft self-test, mc=false."
qwen-2.5-14b-instruct,mmlu_pro,63.2,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4 paper Table 1 baseline (qwen-2.5-14b-instruct column): mmlu_pro=63.2. Third-party Microsoft self-test, mc=false."
llama-3.3-70b-instruct,mmlu_pro,64.4,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4 paper Table 1 baseline (llama-3.3-70b-instruct column): mmlu_pro=64.4. Third-party Microsoft self-test, mc=false."
phi-3-14b,humaneval_plus,69.2,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4 paper Table 1 baseline (phi-3-14b column): humaneval_plus=69.2. Third-party Microsoft self-test, mc=false."
qwen-2.5-14b-instruct,humaneval_plus,79.1,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4 paper Table 1 baseline (qwen-2.5-14b-instruct column): humaneval_plus=79.1. Third-party Microsoft self-test, mc=false."
llama-3.3-70b-instruct,humaneval_plus,77.9,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4 paper Table 1 baseline (llama-3.3-70b-instruct column): humaneval_plus=77.9. Third-party Microsoft self-test, mc=false."
phi-3-14b,arena_hard,45.8,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4 paper Table 1 baseline (phi-3-14b column): arena_hard=45.8. Third-party Microsoft self-test, mc=false."
qwen-2.5-14b-instruct,arena_hard,70.2,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4 paper Table 1 baseline (qwen-2.5-14b-instruct column): arena_hard=70.2. Third-party Microsoft self-test, mc=false."
llama-3.3-70b-instruct,arena_hard,65.5,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4 paper Table 1 baseline (llama-3.3-70b-instruct column): arena_hard=65.5. Third-party Microsoft self-test, mc=false."
phi-3-14b,livebench,28.1,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4 paper Table 1 baseline (phi-3-14b column): livebench=28.1. Third-party Microsoft self-test, mc=false."
qwen-2.5-14b-instruct,livebench,46.6,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4 paper Table 1 baseline (qwen-2.5-14b-instruct column): livebench=46.6. Third-party Microsoft self-test, mc=false."
llama-3.3-70b-instruct,livebench,57.6,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4 paper Table 1 baseline (llama-3.3-70b-instruct column): livebench=57.6. Third-party Microsoft self-test, mc=false."
phi-3-14b,ifeval,57.9,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4 paper Table 1 baseline (phi-3-14b column): ifeval=57.9. Third-party Microsoft self-test, mc=false."
qwen-2.5-14b-instruct,ifeval,78.7,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4 paper Table 1 baseline (qwen-2.5-14b-instruct column): ifeval=78.7. Third-party Microsoft self-test, mc=false."
llama-3.3-70b-instruct,ifeval,89.3,https://arxiv.org/abs/2412.08905,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4 paper Table 1 baseline (llama-3.3-70b-instruct column): ifeval=89.3. Third-party Microsoft self-test, mc=false."
openthinker2-32b,aime_2024,58.0,https://arxiv.org/abs/2504.21318,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4-reasoning paper Table 1 baseline (openthinker2-32b column): aime_2024=58.0. Third-party Microsoft self-test, mc=false."
exaone-deep-32b,aime_2024,72.1,https://arxiv.org/abs/2504.21318,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4-reasoning paper Table 1 baseline (exaone-deep-32b column): aime_2024=72.1. Third-party Microsoft self-test, mc=false."
o1-mini,aime_2024,63.6,https://arxiv.org/abs/2504.21318,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4-reasoning paper Table 1 baseline (o1-mini column): aime_2024=63.6. Third-party Microsoft self-test, mc=false."
openthinker2-32b,aime_2025,58.0,https://arxiv.org/abs/2504.21318,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4-reasoning paper Table 1 baseline (openthinker2-32b column): aime_2025=58.0. Third-party Microsoft self-test, mc=false."
exaone-deep-32b,aime_2025,65.8,https://arxiv.org/abs/2504.21318,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4-reasoning paper Table 1 baseline (exaone-deep-32b column): aime_2025=65.8. Third-party Microsoft self-test, mc=false."
o1-mini,aime_2025,54.8,https://arxiv.org/abs/2504.21318,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4-reasoning paper Table 1 baseline (o1-mini column): aime_2025=54.8. Third-party Microsoft self-test, mc=false."
o1-mini,hmmt_feb_2025,38.0,https://arxiv.org/abs/2504.21318,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4-reasoning paper Table 1 baseline (o1-mini column): hmmt_2025=38.0. Third-party Microsoft self-test, mc=false."
o1-mini,omnimath,60.5,https://arxiv.org/abs/2504.21318,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4-reasoning paper Table 1 baseline (o1-mini column): omnimath=60.5. Third-party Microsoft self-test, mc=false."
openthinker2-32b,gpqa_diamond,64.1,https://arxiv.org/abs/2504.21318,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4-reasoning paper Table 1 baseline (openthinker2-32b column): gpqa_diamond=64.1. Third-party Microsoft self-test, mc=false."
exaone-deep-32b,gpqa_diamond,66.1,https://arxiv.org/abs/2504.21318,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4-reasoning paper Table 1 baseline (exaone-deep-32b column): gpqa_diamond=66.1. Third-party Microsoft self-test, mc=false."
o1-mini,gpqa_diamond,60.0,https://arxiv.org/abs/2504.21318,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4-reasoning paper Table 1 baseline (o1-mini column): gpqa_diamond=60.0. Third-party Microsoft self-test, mc=false."
exaone-deep-32b,livecodebench,59.5,https://arxiv.org/abs/2504.21318,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4-reasoning paper Table 1 baseline (exaone-deep-32b column): livecodebench=59.5. Third-party Microsoft self-test, mc=false."
o1-mini,livecodebench,53.8,https://arxiv.org/abs/2504.21318,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4-reasoning paper Table 1 baseline (o1-mini column): livecodebench=53.8. Third-party Microsoft self-test, mc=false."
o1-mini,codeforces_rating,1650.0,https://arxiv.org/abs/2504.21318,"{""effort"":""unknown"",""harness"":""Microsoft Phi internal eval"",""judge"":""rule-based"",""mode"":""unknown"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",False,tech_report,verified,"Phi-4-reasoning paper Table 1 baseline (o1-mini column): codeforces_rating=1650. Third-party Microsoft self-test, mc=false."
gpt-3.5-turbo,alpacaeval_2,38.7,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gpt-3.5-turbo column): alpacaeval_2=38.7. Third-party Allen AI self-test, mc=false."
gpt-3.5-turbo,bigbench_hard,66.6,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gpt-3.5-turbo column): bigbench_hard=66.6. Third-party Allen AI self-test, mc=false."
gpt-3.5-turbo,drop,70.2,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gpt-3.5-turbo column): drop=70.2. Third-party Allen AI self-test, mc=false."
gpt-3.5-turbo,gsm8k,74.3,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gpt-3.5-turbo column): gsm8k=74.3. Third-party Allen AI self-test, mc=false."
gpt-3.5-turbo,ifeval,66.9,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gpt-3.5-turbo column): ifeval=66.9. Third-party Allen AI self-test, mc=false."
gpt-3.5-turbo,math,41.2,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gpt-3.5-turbo column): math=41.2. Third-party Allen AI self-test, mc=false."
gpt-3.5-turbo,mmlu,70.2,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gpt-3.5-turbo column): mmlu=70.2. Third-party Allen AI self-test, mc=false."
gpt-3.5-turbo,safety,69.1,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gpt-3.5-turbo column): safety=69.1. Third-party Allen AI self-test, mc=false."
gpt-3.5-turbo,popqa,45.0,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gpt-3.5-turbo column): popqa=45.0. Third-party Allen AI self-test, mc=false."
gpt-3.5-turbo,truthfulqa,62.9,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gpt-3.5-turbo column): truthfulqa=62.9. Third-party Allen AI self-test, mc=false."
gpt-4o-mini,alpacaeval_2,49.7,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gpt-4o-mini column): alpacaeval_2=49.7. Third-party Allen AI self-test, mc=false."
gpt-4o-mini,bigbench_hard,65.9,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gpt-4o-mini column): bigbench_hard=65.9. Third-party Allen AI self-test, mc=false."
gpt-4o-mini,gsm8k,83.0,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gpt-4o-mini column): gsm8k=83.0. Third-party Allen AI self-test, mc=false."
gpt-4o-mini,math,67.9,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gpt-4o-mini column): math=67.9. Third-party Allen AI self-test, mc=false."
gpt-4o-mini,safety,84.9,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gpt-4o-mini column): safety=84.9. Third-party Allen AI self-test, mc=false."
gpt-4o-mini,popqa,39.0,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gpt-4o-mini column): popqa=39.0. Third-party Allen AI self-test, mc=false."
gpt-4o-mini,truthfulqa,64.8,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gpt-4o-mini column): truthfulqa=64.8. Third-party Allen AI self-test, mc=false."
gemma-3-1b,alpacaeval_2,20.4,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gemma-3-1b column): alpacaeval_2=20.4. Third-party Allen AI self-test, mc=false."
gemma-3-1b,bigbench_hard,39.4,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gemma-3-1b column): bigbench_hard=39.4. Third-party Allen AI self-test, mc=false."
gemma-3-1b,drop,25.1,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gemma-3-1b column): drop=25.1. Third-party Allen AI self-test, mc=false."
gemma-3-1b,gsm8k,35.0,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gemma-3-1b column): gsm8k=35.0. Third-party Allen AI self-test, mc=false."
gemma-3-1b,ifeval,60.6,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gemma-3-1b column): ifeval=60.6. Third-party Allen AI self-test, mc=false."
gemma-3-1b,mmlu,38.9,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gemma-3-1b column): mmlu=38.9. Third-party Allen AI self-test, mc=false."
gemma-3-1b,safety,70.2,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gemma-3-1b column): safety=70.2. Third-party Allen AI self-test, mc=false."
gemma-3-1b,popqa,9.6,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gemma-3-1b column): popqa=9.6. Third-party Allen AI self-test, mc=false."
gemma-3-1b,truthfulqa,43.8,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gemma-3-1b column): truthfulqa=43.8. Third-party Allen AI self-test, mc=false."
llama-3.2-1b,alpacaeval_2,10.1,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (llama-3.2-1b column): alpacaeval_2=10.1. Third-party Allen AI self-test, mc=false."
llama-3.2-1b,bigbench_hard,40.2,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (llama-3.2-1b column): bigbench_hard=40.2. Third-party Allen AI self-test, mc=false."
llama-3.2-1b,drop,32.2,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (llama-3.2-1b column): drop=32.2. Third-party Allen AI self-test, mc=false."
llama-3.2-1b,gsm8k,45.4,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (llama-3.2-1b column): gsm8k=45.4. Third-party Allen AI self-test, mc=false."
llama-3.2-1b,ifeval,54.0,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (llama-3.2-1b column): ifeval=54.0. Third-party Allen AI self-test, mc=false."
llama-3.2-1b,math,21.6,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (llama-3.2-1b column): math=21.6. Third-party Allen AI self-test, mc=false."
llama-3.2-1b,mmlu,46.7,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (llama-3.2-1b column): mmlu=46.7. Third-party Allen AI self-test, mc=false."
llama-3.2-1b,safety,87.2,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (llama-3.2-1b column): safety=87.2. Third-party Allen AI self-test, mc=false."
llama-3.2-1b,popqa,13.8,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (llama-3.2-1b column): popqa=13.8. Third-party Allen AI self-test, mc=false."
llama-3.2-1b,truthfulqa,41.5,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (llama-3.2-1b column): truthfulqa=41.5. Third-party Allen AI self-test, mc=false."
qwen-2.5-1.5b,alpacaeval_2,7.4,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-1.5b column): alpacaeval_2=7.4. Third-party Allen AI self-test, mc=false."
qwen-2.5-1.5b,bigbench_hard,45.8,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-1.5b column): bigbench_hard=45.8. Third-party Allen AI self-test, mc=false."
qwen-2.5-1.5b,drop,13.4,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-1.5b column): drop=13.4. Third-party Allen AI self-test, mc=false."
qwen-2.5-1.5b,gsm8k,66.2,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-1.5b column): gsm8k=66.2. Third-party Allen AI self-test, mc=false."
qwen-2.5-1.5b,ifeval,44.2,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-1.5b column): ifeval=44.2. Third-party Allen AI self-test, mc=false."
qwen-2.5-1.5b,math,40.6,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-1.5b column): math=40.6. Third-party Allen AI self-test, mc=false."
qwen-2.5-1.5b,mmlu,59.7,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-1.5b column): mmlu=59.7. Third-party Allen AI self-test, mc=false."
qwen-2.5-1.5b,safety,77.6,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-1.5b column): safety=77.6. Third-party Allen AI self-test, mc=false."
qwen-2.5-1.5b,popqa,15.5,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-1.5b column): popqa=15.5. Third-party Allen AI self-test, mc=false."
qwen-2.5-1.5b,truthfulqa,46.5,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-1.5b column): truthfulqa=46.5. Third-party Allen AI self-test, mc=false."
ministral-8b,alpacaeval_2,31.4,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (ministral-8b column): alpacaeval_2=31.4. Third-party Allen AI self-test, mc=false."
ministral-8b,bigbench_hard,70.8,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (ministral-8b column): bigbench_hard=70.8. Third-party Allen AI self-test, mc=false."
ministral-8b,drop,56.2,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (ministral-8b column): drop=56.2. Third-party Allen AI self-test, mc=false."
ministral-8b,gsm8k,80.0,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (ministral-8b column): gsm8k=80.0. Third-party Allen AI self-test, mc=false."
ministral-8b,ifeval,56.4,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (ministral-8b column): ifeval=56.4. Third-party Allen AI self-test, mc=false."
ministral-8b,math,40.0,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (ministral-8b column): math=40.0. Third-party Allen AI self-test, mc=false."
ministral-8b,mmlu,68.5,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (ministral-8b column): mmlu=68.5. Third-party Allen AI self-test, mc=false."
ministral-8b,safety,56.2,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (ministral-8b column): safety=56.2. Third-party Allen AI self-test, mc=false."
ministral-8b,popqa,20.2,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (ministral-8b column): popqa=20.2. Third-party Allen AI self-test, mc=false."
ministral-8b,truthfulqa,55.5,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (ministral-8b column): truthfulqa=55.5. Third-party Allen AI self-test, mc=false."
llama-3.1-8b,alpacaeval_2,25.8,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (llama-3.1-8b column): alpacaeval_2=25.8. Third-party Allen AI self-test, mc=false."
llama-3.1-8b,bigbench_hard,71.9,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (llama-3.1-8b column): bigbench_hard=71.9. Third-party Allen AI self-test, mc=false."
llama-3.1-8b,drop,61.7,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (llama-3.1-8b column): drop=61.7. Third-party Allen AI self-test, mc=false."
llama-3.1-8b,gsm8k,83.4,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (llama-3.1-8b column): gsm8k=83.4. Third-party Allen AI self-test, mc=false."
llama-3.1-8b,ifeval,80.6,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (llama-3.1-8b column): ifeval=80.6. Third-party Allen AI self-test, mc=false."
llama-3.1-8b,math,42.5,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (llama-3.1-8b column): math=42.5. Third-party Allen AI self-test, mc=false."
llama-3.1-8b,mmlu,71.3,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (llama-3.1-8b column): mmlu=71.3. Third-party Allen AI self-test, mc=false."
llama-3.1-8b,safety,70.2,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (llama-3.1-8b column): safety=70.2. Third-party Allen AI self-test, mc=false."
llama-3.1-8b,popqa,28.4,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (llama-3.1-8b column): popqa=28.4. Third-party Allen AI self-test, mc=false."
llama-3.1-8b,truthfulqa,55.1,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (llama-3.1-8b column): truthfulqa=55.1. Third-party Allen AI self-test, mc=false."
tulu-3-8b,alpacaeval_2,34.0,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (tulu-3-8b column): alpacaeval_2=34.0. Third-party Allen AI self-test, mc=false."
tulu-3-8b,bigbench_hard,69.0,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (tulu-3-8b column): bigbench_hard=69.0. Third-party Allen AI self-test, mc=false."
tulu-3-8b,drop,62.6,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (tulu-3-8b column): drop=62.6. Third-party Allen AI self-test, mc=false."
tulu-3-8b,gsm8k,87.6,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (tulu-3-8b column): gsm8k=87.6. Third-party Allen AI self-test, mc=false."
tulu-3-8b,ifeval,82.4,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (tulu-3-8b column): ifeval=82.4. Third-party Allen AI self-test, mc=false."
tulu-3-8b,math,43.7,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (tulu-3-8b column): math=43.7. Third-party Allen AI self-test, mc=false."
tulu-3-8b,mmlu,68.2,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (tulu-3-8b column): mmlu=68.2. Third-party Allen AI self-test, mc=false."
tulu-3-8b,safety,75.4,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (tulu-3-8b column): safety=75.4. Third-party Allen AI self-test, mc=false."
tulu-3-8b,popqa,29.1,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (tulu-3-8b column): popqa=29.1. Third-party Allen AI self-test, mc=false."
tulu-3-8b,truthfulqa,55.0,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (tulu-3-8b column): truthfulqa=55.0. Third-party Allen AI self-test, mc=false."
qwen-2.5-7b,alpacaeval_2,29.7,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-7b column): alpacaeval_2=29.7. Third-party Allen AI self-test, mc=false."
qwen-2.5-7b,bigbench_hard,70.2,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-7b column): bigbench_hard=70.2. Third-party Allen AI self-test, mc=false."
qwen-2.5-7b,drop,54.4,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-7b column): drop=54.4. Third-party Allen AI self-test, mc=false."
qwen-2.5-7b,gsm8k,83.8,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-7b column): gsm8k=83.8. Third-party Allen AI self-test, mc=false."
qwen-2.5-7b,ifeval,74.7,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-7b column): ifeval=74.7. Third-party Allen AI self-test, mc=false."
qwen-2.5-7b,math,69.9,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-7b column): math=69.9. Third-party Allen AI self-test, mc=false."
qwen-2.5-7b,mmlu,76.6,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-7b column): mmlu=76.6. Third-party Allen AI self-test, mc=false."
qwen-2.5-7b,safety,75.0,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-7b column): safety=75.0. Third-party Allen AI self-test, mc=false."
qwen-2.5-7b,popqa,18.1,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-7b column): popqa=18.1. Third-party Allen AI self-test, mc=false."
qwen-2.5-7b,truthfulqa,63.1,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-7b column): truthfulqa=63.1. Third-party Allen AI self-test, mc=false."
gemma-2-9b,alpacaeval_2,43.7,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gemma-2-9b column): alpacaeval_2=43.7. Third-party Allen AI self-test, mc=false."
gemma-2-9b,bigbench_hard,64.9,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gemma-2-9b column): bigbench_hard=64.9. Third-party Allen AI self-test, mc=false."
gemma-2-9b,drop,58.8,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gemma-2-9b column): drop=58.8. Third-party Allen AI self-test, mc=false."
gemma-2-9b,gsm8k,79.7,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gemma-2-9b column): gsm8k=79.7. Third-party Allen AI self-test, mc=false."
gemma-2-9b,ifeval,69.9,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gemma-2-9b column): ifeval=69.9. Third-party Allen AI self-test, mc=false."
gemma-2-9b,mmlu,69.1,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gemma-2-9b column): mmlu=69.1. Third-party Allen AI self-test, mc=false."
gemma-2-9b,safety,75.5,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gemma-2-9b column): safety=75.5. Third-party Allen AI self-test, mc=false."
gemma-2-9b,popqa,28.3,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gemma-2-9b column): popqa=28.3. Third-party Allen AI self-test, mc=false."
gemma-2-9b,truthfulqa,61.4,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gemma-2-9b column): truthfulqa=61.4. Third-party Allen AI self-test, mc=false."
qwen-2.5-14b,alpacaeval_2,34.6,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-14b column): alpacaeval_2=34.6. Third-party Allen AI self-test, mc=false."
qwen-2.5-14b,bigbench_hard,78.4,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-14b column): bigbench_hard=78.4. Third-party Allen AI self-test, mc=false."
qwen-2.5-14b,drop,50.5,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-14b column): drop=50.5. Third-party Allen AI self-test, mc=false."
qwen-2.5-14b,gsm8k,83.9,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-14b column): gsm8k=83.9. Third-party Allen AI self-test, mc=false."
qwen-2.5-14b,ifeval,82.4,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-14b column): ifeval=82.4. Third-party Allen AI self-test, mc=false."
qwen-2.5-14b,math,70.6,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-14b column): math=70.6. Third-party Allen AI self-test, mc=false."
qwen-2.5-14b,mmlu,81.1,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-14b column): mmlu=81.1. Third-party Allen AI self-test, mc=false."
qwen-2.5-14b,safety,79.3,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-14b column): safety=79.3. Third-party Allen AI self-test, mc=false."
qwen-2.5-14b,popqa,21.1,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-14b column): popqa=21.1. Third-party Allen AI self-test, mc=false."
qwen-2.5-14b,truthfulqa,70.8,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-14b column): truthfulqa=70.8. Third-party Allen AI self-test, mc=false."
gemma-2-27b,alpacaeval_2,49.0,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gemma-2-27b column): alpacaeval_2=49.0. Third-party Allen AI self-test, mc=false."
gemma-2-27b,bigbench_hard,72.7,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gemma-2-27b column): bigbench_hard=72.7. Third-party Allen AI self-test, mc=false."
gemma-2-27b,drop,67.5,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gemma-2-27b column): drop=67.5. Third-party Allen AI self-test, mc=false."
gemma-2-27b,gsm8k,80.7,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gemma-2-27b column): gsm8k=80.7. Third-party Allen AI self-test, mc=false."
gemma-2-27b,ifeval,63.2,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gemma-2-27b column): ifeval=63.2. Third-party Allen AI self-test, mc=false."
gemma-2-27b,mmlu,70.7,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gemma-2-27b column): mmlu=70.7. Third-party Allen AI self-test, mc=false."
gemma-2-27b,safety,75.9,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gemma-2-27b column): safety=75.9. Third-party Allen AI self-test, mc=false."
gemma-2-27b,popqa,33.9,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gemma-2-27b column): popqa=33.9. Third-party Allen AI self-test, mc=false."
gemma-2-27b,truthfulqa,64.6,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gemma-2-27b column): truthfulqa=64.6. Third-party Allen AI self-test, mc=false."
qwen-2.5-32b,alpacaeval_2,39.1,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-32b column): alpacaeval_2=39.1. Third-party Allen AI self-test, mc=false."
qwen-2.5-32b,bigbench_hard,82.3,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-32b column): bigbench_hard=82.3. Third-party Allen AI self-test, mc=false."
qwen-2.5-32b,drop,48.3,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-32b column): drop=48.3. Third-party Allen AI self-test, mc=false."
qwen-2.5-32b,gsm8k,87.5,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-32b column): gsm8k=87.5. Third-party Allen AI self-test, mc=false."
qwen-2.5-32b,ifeval,82.4,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-32b column): ifeval=82.4. Third-party Allen AI self-test, mc=false."
qwen-2.5-32b,math,77.9,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-32b column): math=77.9. Third-party Allen AI self-test, mc=false."
qwen-2.5-32b,mmlu,84.7,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-32b column): mmlu=84.7. Third-party Allen AI self-test, mc=false."
qwen-2.5-32b,safety,82.4,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-32b column): safety=82.4. Third-party Allen AI self-test, mc=false."
qwen-2.5-32b,popqa,26.1,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-32b column): popqa=26.1. Third-party Allen AI self-test, mc=false."
qwen-2.5-32b,truthfulqa,70.6,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen-2.5-32b column): truthfulqa=70.6. Third-party Allen AI self-test, mc=false."
mistral-small-24b,alpacaeval_2,43.2,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (mistral-small-24b column): alpacaeval_2=43.2. Third-party Allen AI self-test, mc=false."
mistral-small-24b,bigbench_hard,80.1,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (mistral-small-24b column): bigbench_hard=80.1. Third-party Allen AI self-test, mc=false."
mistral-small-24b,drop,78.5,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (mistral-small-24b column): drop=78.5. Third-party Allen AI self-test, mc=false."
mistral-small-24b,gsm8k,87.2,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (mistral-small-24b column): gsm8k=87.2. Third-party Allen AI self-test, mc=false."
mistral-small-24b,ifeval,77.3,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (mistral-small-24b column): ifeval=77.3. Third-party Allen AI self-test, mc=false."
mistral-small-24b,math,65.9,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (mistral-small-24b column): math=65.9. Third-party Allen AI self-test, mc=false."
mistral-small-24b,mmlu,83.7,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (mistral-small-24b column): mmlu=83.7. Third-party Allen AI self-test, mc=false."
mistral-small-24b,safety,66.5,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (mistral-small-24b column): safety=66.5. Third-party Allen AI self-test, mc=false."
mistral-small-24b,popqa,24.4,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (mistral-small-24b column): popqa=24.4. Third-party Allen AI self-test, mc=false."
mistral-small-24b,truthfulqa,68.1,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (mistral-small-24b column): truthfulqa=68.1. Third-party Allen AI self-test, mc=false."
qwq-32b,alpacaeval_2,82.4,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwq-32b column): alpacaeval_2=82.4. Third-party Allen AI self-test, mc=false."
qwq-32b,bigbench_hard,89.6,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwq-32b column): bigbench_hard=89.6. Third-party Allen AI self-test, mc=false."
qwq-32b,drop,54.7,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwq-32b column): drop=54.7. Third-party Allen AI self-test, mc=false."
qwq-32b,gsm8k,95.5,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwq-32b column): gsm8k=95.5. Third-party Allen AI self-test, mc=false."
qwq-32b,math,98.1,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwq-32b column): math=98.1. Third-party Allen AI self-test, mc=false."
qwq-32b,safety,69.9,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwq-32b column): safety=69.9. Third-party Allen AI self-test, mc=false."
gemma-3-27b,alpacaeval_2,63.4,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gemma-3-27b column): alpacaeval_2=63.4. Third-party Allen AI self-test, mc=false."
gemma-3-27b,bigbench_hard,83.7,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gemma-3-27b column): bigbench_hard=83.7. Third-party Allen AI self-test, mc=false."
gemma-3-27b,drop,69.2,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gemma-3-27b column): drop=69.2. Third-party Allen AI self-test, mc=false."
gemma-3-27b,safety,69.1,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gemma-3-27b column): safety=69.1. Third-party Allen AI self-test, mc=false."
gemma-3-27b,popqa,30.9,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gemma-3-27b column): popqa=30.9. Third-party Allen AI self-test, mc=false."
gemma-3-27b,truthfulqa,63.9,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (gemma-3-27b column): truthfulqa=63.9. Third-party Allen AI self-test, mc=false."
qwen2.5-72b-instruct,alpacaeval_2,47.7,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen2.5-72b-instruct column): alpacaeval_2=47.7. Third-party Allen AI self-test, mc=false."
qwen2.5-72b-instruct,bigbench_hard,80.4,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen2.5-72b-instruct column): bigbench_hard=80.4. Third-party Allen AI self-test, mc=false."
qwen2.5-72b-instruct,gsm8k,89.5,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen2.5-72b-instruct column): gsm8k=89.5. Third-party Allen AI self-test, mc=false."
qwen2.5-72b-instruct,math,75.9,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen2.5-72b-instruct column): math=75.9. Third-party Allen AI self-test, mc=false."
qwen2.5-72b-instruct,safety,87.0,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen2.5-72b-instruct column): safety=87.0. Third-party Allen AI self-test, mc=false."
qwen2.5-72b-instruct,popqa,30.6,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen2.5-72b-instruct column): popqa=30.6. Third-party Allen AI self-test, mc=false."
qwen2.5-72b-instruct,truthfulqa,69.9,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (qwen2.5-72b-instruct column): truthfulqa=69.9. Third-party Allen AI self-test, mc=false."
llama-3.1-70b,alpacaeval_2,32.9,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (llama-3.1-70b column): alpacaeval_2=32.9. Third-party Allen AI self-test, mc=false."
llama-3.1-70b,bigbench_hard,83.0,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (llama-3.1-70b column): bigbench_hard=83.0. Third-party Allen AI self-test, mc=false."
llama-3.1-70b,drop,77.0,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (llama-3.1-70b column): drop=77.0. Third-party Allen AI self-test, mc=false."
llama-3.1-70b,gsm8k,94.5,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (llama-3.1-70b column): gsm8k=94.5. Third-party Allen AI self-test, mc=false."
llama-3.1-70b,ifeval,88.0,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (llama-3.1-70b column): ifeval=88.0. Third-party Allen AI self-test, mc=false."
llama-3.1-70b,math,56.2,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (llama-3.1-70b column): math=56.2. Third-party Allen AI self-test, mc=false."
llama-3.1-70b,mmlu,85.2,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (llama-3.1-70b column): mmlu=85.2. Third-party Allen AI self-test, mc=false."
llama-3.1-70b,safety,76.4,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (llama-3.1-70b column): safety=76.4. Third-party Allen AI self-test, mc=false."
llama-3.1-70b,popqa,46.5,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (llama-3.1-70b column): popqa=46.5. Third-party Allen AI self-test, mc=false."
llama-3.1-70b,truthfulqa,66.8,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (llama-3.1-70b column): truthfulqa=66.8. Third-party Allen AI self-test, mc=false."
llama-3.3-70b-instruct,alpacaeval_2,36.5,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (llama-3.3-70b-instruct column): alpacaeval_2=36.5. Third-party Allen AI self-test, mc=false."
llama-3.3-70b-instruct,bigbench_hard,85.8,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (llama-3.3-70b-instruct column): bigbench_hard=85.8. Third-party Allen AI self-test, mc=false."
llama-3.3-70b-instruct,gsm8k,93.6,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (llama-3.3-70b-instruct column): gsm8k=93.6. Third-party Allen AI self-test, mc=false."
llama-3.3-70b-instruct,math,71.8,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (llama-3.3-70b-instruct column): math=71.8. Third-party Allen AI self-test, mc=false."
llama-3.3-70b-instruct,safety,70.4,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (llama-3.3-70b-instruct column): safety=70.4. Third-party Allen AI self-test, mc=false."
llama-3.3-70b-instruct,popqa,48.2,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (llama-3.3-70b-instruct column): popqa=48.2. Third-party Allen AI self-test, mc=false."
llama-3.3-70b-instruct,truthfulqa,66.1,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"OLMo 2 paper Table 7 Instruct (llama-3.3-70b-instruct column): truthfulqa=66.1. Third-party Allen AI self-test, mc=false."
olmo-1b,alpacaeval_2,2.4,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-1b column): alpacaeval_2=2.4.
olmo-1b,bigbench_hard,29.9,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-1b column): bigbench_hard=29.9.
olmo-1b,drop,27.9,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-1b column): drop=27.9.
olmo-1b,gsm8k,10.8,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-1b column): gsm8k=10.8.
olmo-1b,ifeval,25.3,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-1b column): ifeval=25.3.
olmo-1b,math,2.2,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-1b column): math=2.2.
olmo-1b,mmlu,36.6,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-1b column): mmlu=36.6.
olmo-1b,safety,52.0,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-1b column): safety=52.0.
olmo-1b,popqa,12.1,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-1b column): popqa=12.1.
olmo-1b,truthfulqa,44.3,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-1b column): truthfulqa=44.3.
smollm2-1.7b,alpacaeval_2,5.8,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (smollm2-1.7b column): alpacaeval_2=5.8.
smollm2-1.7b,bigbench_hard,39.8,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (smollm2-1.7b column): bigbench_hard=39.8.
smollm2-1.7b,drop,30.9,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (smollm2-1.7b column): drop=30.9.
smollm2-1.7b,gsm8k,45.3,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (smollm2-1.7b column): gsm8k=45.3.
smollm2-1.7b,ifeval,51.6,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (smollm2-1.7b column): ifeval=51.6.
smollm2-1.7b,math,20.3,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (smollm2-1.7b column): math=20.3.
smollm2-1.7b,mmlu,34.3,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (smollm2-1.7b column): mmlu=34.3.
smollm2-1.7b,safety,52.4,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (smollm2-1.7b column): safety=52.4.
smollm2-1.7b,popqa,16.4,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (smollm2-1.7b column): popqa=16.4.
smollm2-1.7b,truthfulqa,45.3,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (smollm2-1.7b column): truthfulqa=45.3.
olmo-7b-0424,alpacaeval_2,8.5,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-7b-0424 column): alpacaeval_2=8.5.
olmo-7b-0424,bigbench_hard,34.4,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-7b-0424 column): bigbench_hard=34.4.
olmo-7b-0424,drop,47.9,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-7b-0424 column): drop=47.9.
olmo-7b-0424,gsm8k,23.2,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-7b-0424 column): gsm8k=23.2.
olmo-7b-0424,ifeval,39.2,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-7b-0424 column): ifeval=39.2.
olmo-7b-0424,math,5.2,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-7b-0424 column): math=5.2.
olmo-7b-0424,mmlu,48.9,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-7b-0424 column): mmlu=48.9.
olmo-7b-0424,safety,49.3,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-7b-0424 column): safety=49.3.
olmo-7b-0424,popqa,18.9,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-7b-0424 column): popqa=18.9.
olmo-7b-0424,truthfulqa,55.2,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-7b-0424 column): truthfulqa=55.2.
olmo-2-1b,alpacaeval_2,9.1,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-2-1b column): alpacaeval_2=9.1.
olmo-2-1b,bigbench_hard,35.0,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-2-1b column): bigbench_hard=35.0.
olmo-2-1b,drop,34.6,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-2-1b column): drop=34.6.
olmo-2-1b,gsm8k,68.3,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-2-1b column): gsm8k=68.3.
olmo-2-1b,ifeval,70.1,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-2-1b column): ifeval=70.1.
olmo-2-1b,math,20.7,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-2-1b column): math=20.7.
olmo-2-1b,mmlu,40.0,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-2-1b column): mmlu=40.0.
olmo-2-1b,safety,87.6,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-2-1b column): safety=87.6.
olmo-2-1b,popqa,12.9,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-2-1b column): popqa=12.9.
olmo-2-1b,truthfulqa,48.7,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-2-1b column): truthfulqa=48.7.
olmo-2-7b,alpacaeval_2,29.1,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-2-7b column): alpacaeval_2=29.1.
olmo-2-7b,bigbench_hard,51.4,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-2-7b column): bigbench_hard=51.4.
olmo-2-7b,drop,60.5,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-2-7b column): drop=60.5.
olmo-2-7b,gsm8k,85.1,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-2-7b column): gsm8k=85.1.
olmo-2-7b,ifeval,72.3,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-2-7b column): ifeval=72.3.
olmo-2-7b,math,32.5,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-2-7b column): math=32.5.
olmo-2-7b,mmlu,61.3,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-2-7b column): mmlu=61.3.
olmo-2-7b,safety,93.3,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-2-7b column): safety=93.3.
olmo-2-7b,popqa,23.2,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-2-7b column): popqa=23.2.
olmo-2-7b,truthfulqa,56.5,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-2-7b column): truthfulqa=56.5.
olmo-2-13b,alpacaeval_2,39.5,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-2-13b column): alpacaeval_2=39.5.
olmo-2-13b,bigbench_hard,63.0,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-2-13b column): bigbench_hard=63.0.
olmo-2-13b,drop,71.5,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-2-13b column): drop=71.5.
olmo-2-13b,gsm8k,87.4,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-2-13b column): gsm8k=87.4.
olmo-2-13b,math,39.2,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-2-13b column): math=39.2.
olmo-2-13b,safety,89.7,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-2-13b column): safety=89.7.
olmo-2-13b,popqa,28.8,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-2-13b column): popqa=28.8.
olmo-2-13b,truthfulqa,64.3,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-2-13b column): truthfulqa=64.3.
olmo-2-32b,alpacaeval_2,42.8,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-2-32b column): alpacaeval_2=42.8.
olmo-2-32b,bigbench_hard,70.6,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-2-32b column): bigbench_hard=70.6.
olmo-2-32b,drop,78.0,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-2-32b column): drop=78.0.
olmo-2-32b,gsm8k,87.6,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-2-32b column): gsm8k=87.6.
olmo-2-32b,ifeval,85.6,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-2-32b column): ifeval=85.6.
olmo-2-32b,math,49.7,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-2-32b column): math=49.7.
olmo-2-32b,mmlu,77.3,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-2-32b column): mmlu=77.3.
olmo-2-32b,safety,85.9,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-2-32b column): safety=85.9.
olmo-2-32b,popqa,37.5,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-2-32b column): popqa=37.5.
olmo-2-32b,truthfulqa,73.2,https://arxiv.org/abs/2501.00656,"{""effort"":""n/a"",""harness"":""OLMES"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,OLMo 2 paper Table 7 Instruct (olmo-2-32b column): truthfulqa=73.2.
lfm2.5-1.2b-thinking,ifbench,44.85,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""default"",""harness"":""official (LFM eval)"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":0.05,""tools"":""none""}",True,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: lfm2.5-1.2b-thinking=44.85.
lfm2.5-1.2b-thinking,multi_if,69.33,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""default"",""harness"":""official (LFM eval)"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":0.05,""tools"":""none""}",True,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: lfm2.5-1.2b-thinking=69.33.
lfm2.5-1.2b-thinking,gsm8k,85.6,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""default"",""harness"":""official (LFM eval)"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":0.05,""tools"":""none""}",True,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: lfm2.5-1.2b-thinking=85.6.
lfm2.5-1.2b-thinking,bfcl_v3,56.97,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""default"",""harness"":""official (LFM eval)"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":0.05,""tools"":""none""}",True,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: lfm2.5-1.2b-thinking=56.97.
qwen3-1.7b,ifbench,25.88,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""default"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: qwen3-1.7b=25.88. (third-party self-test by Liquid AI)
lfm2.5-1.2b-instruct,gpqa_diamond,38.89,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: lfm2.5-1.2b-instruct=38.89. (third-party self-test by Liquid AI)
lfm2.5-1.2b-instruct,mmlu_pro,44.35,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: lfm2.5-1.2b-instruct=44.35. (third-party self-test by Liquid AI)
lfm2.5-1.2b-instruct,ifeval,86.23,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: lfm2.5-1.2b-instruct=86.23. (third-party self-test by Liquid AI)
lfm2.5-1.2b-instruct,ifbench,47.33,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: lfm2.5-1.2b-instruct=47.33. (third-party self-test by Liquid AI)
lfm2.5-1.2b-instruct,multi_if,60.98,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: lfm2.5-1.2b-instruct=60.98. (third-party self-test by Liquid AI)
lfm2.5-1.2b-instruct,gsm8k,64.52,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: lfm2.5-1.2b-instruct=64.52. (third-party self-test by Liquid AI)
lfm2.5-1.2b-instruct,math_500,63.2,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: lfm2.5-1.2b-instruct=63.2. (third-party self-test by Liquid AI)
lfm2.5-1.2b-instruct,aime_2025,14.0,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: lfm2.5-1.2b-instruct=14.0. (third-party self-test by Liquid AI)
lfm2.5-1.2b-instruct,bfcl_v3,49.12,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: lfm2.5-1.2b-instruct=49.12. (third-party self-test by Liquid AI)
granite-4.0-h-1b,gpqa_diamond,24.34,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: granite-4.0-h-1b=24.34. (third-party self-test by Liquid AI)
granite-4.0-h-1b,mmlu_pro,27.64,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: granite-4.0-h-1b=27.64. (third-party self-test by Liquid AI)
granite-4.0-h-1b,ifeval,80.08,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: granite-4.0-h-1b=80.08. (third-party self-test by Liquid AI)
granite-4.0-h-1b,ifbench,24.93,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: granite-4.0-h-1b=24.93. (third-party self-test by Liquid AI)
granite-4.0-h-1b,multi_if,47.56,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: granite-4.0-h-1b=47.56. (third-party self-test by Liquid AI)
granite-4.0-h-1b,gsm8k,69.6,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: granite-4.0-h-1b=69.6. (third-party self-test by Liquid AI)
granite-4.0-h-1b,math_500,47.2,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: granite-4.0-h-1b=47.2. (third-party self-test by Liquid AI)
granite-4.0-h-1b,aime_2025,1.0,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: granite-4.0-h-1b=1.0. (third-party self-test by Liquid AI)
granite-4.0-h-1b,bfcl_v3,50.69,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: granite-4.0-h-1b=50.69. (third-party self-test by Liquid AI)
granite-4.0-1b,gpqa_diamond,24.24,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: granite-4.0-1b=24.24. (third-party self-test by Liquid AI)
granite-4.0-1b,mmlu_pro,33.53,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: granite-4.0-1b=33.53. (third-party self-test by Liquid AI)
granite-4.0-1b,ifeval,79.61,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: granite-4.0-1b=79.61. (third-party self-test by Liquid AI)
granite-4.0-1b,ifbench,21.0,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: granite-4.0-1b=21.0. (third-party self-test by Liquid AI)
granite-4.0-1b,multi_if,43.65,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: granite-4.0-1b=43.65. (third-party self-test by Liquid AI)
granite-4.0-1b,gsm8k,73.42,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: granite-4.0-1b=73.42. (third-party self-test by Liquid AI)
granite-4.0-1b,math_500,44.8,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: granite-4.0-1b=44.8. (third-party self-test by Liquid AI)
granite-4.0-1b,aime_2025,3.33,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: granite-4.0-1b=3.33. (third-party self-test by Liquid AI)
granite-4.0-1b,bfcl_v3,52.43,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: granite-4.0-1b=52.43. (third-party self-test by Liquid AI)
gemma-3-1b-it,gpqa_diamond,24.24,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: gemma-3-1b-it=24.24. (third-party self-test by Liquid AI)
gemma-3-1b-it,mmlu_pro,14.04,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: gemma-3-1b-it=14.04. (third-party self-test by Liquid AI)
gemma-3-1b-it,ifeval,63.25,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: gemma-3-1b-it=63.25. (third-party self-test by Liquid AI)
gemma-3-1b-it,ifbench,20.47,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: gemma-3-1b-it=20.47. (third-party self-test by Liquid AI)
gemma-3-1b-it,multi_if,44.31,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: gemma-3-1b-it=44.31. (third-party self-test by Liquid AI)
gemma-3-1b-it,gsm8k,42.15,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: gemma-3-1b-it=42.15. (third-party self-test by Liquid AI)
gemma-3-1b-it,math_500,45.2,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: gemma-3-1b-it=45.2. (third-party self-test by Liquid AI)
gemma-3-1b-it,aime_2025,1.0,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: gemma-3-1b-it=1.0. (third-party self-test by Liquid AI)
gemma-3-1b-it,bfcl_v3,16.64,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: gemma-3-1b-it=16.64. (third-party self-test by Liquid AI)
llama-3.2-1b-instruct,gpqa_diamond,16.57,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: llama-3.2-1b-instruct=16.57. (third-party self-test by Liquid AI)
llama-3.2-1b-instruct,mmlu_pro,20.8,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: llama-3.2-1b-instruct=20.8. (third-party self-test by Liquid AI)
llama-3.2-1b-instruct,ifeval,52.37,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: llama-3.2-1b-instruct=52.37. (third-party self-test by Liquid AI)
llama-3.2-1b-instruct,ifbench,15.93,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: llama-3.2-1b-instruct=15.93. (third-party self-test by Liquid AI)
llama-3.2-1b-instruct,multi_if,30.16,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: llama-3.2-1b-instruct=30.16. (third-party self-test by Liquid AI)
llama-3.2-1b-instruct,gsm8k,39.04,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: llama-3.2-1b-instruct=39.04. (third-party self-test by Liquid AI)
llama-3.2-1b-instruct,math_500,23.4,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: llama-3.2-1b-instruct=23.4. (third-party self-test by Liquid AI)
llama-3.2-1b-instruct,aime_2025,0.33,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: llama-3.2-1b-instruct=0.33. (third-party self-test by Liquid AI)
llama-3.2-1b-instruct,bfcl_v3,21.44,https://www.liquid.ai/blog/lfm2-5-1-2b-thinking-on-device-reasoning-under-1gb,"{""effort"":""n/a"",""harness"":""official (LFM eval, third-party)"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (mean of 5 runs)"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Liquid AI LFM2.5-1.2B-Thinking blog: llama-3.2-1b-instruct=21.44. (third-party self-test by Liquid AI)
mistral-medium-3,livecodebench_v6,30.3,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",True,official_blog,verified,Mistral Medium 3 blog table: 30.3.
mistral-medium-3,multipl_e_avg,81.4,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",True,official_blog,verified,Mistral Medium 3 blog table: 81.4.
mistral-medium-3,ruler_32k,96.0,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",True,official_blog,verified,Mistral Medium 3 blog table: 96.0.
mistral-medium-3,ruler_128k,90.2,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",True,official_blog,verified,Mistral Medium 3 blog table: 90.2.
mistral-medium-3,mmmu,66.1,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",True,official_blog,verified,Mistral Medium 3 blog table: 66.1.
mistral-medium-3,docvqa,95.3,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",True,official_blog,verified,Mistral Medium 3 blog table: 95.3.
mistral-medium-3,ai2d,93.7,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",True,official_blog,verified,Mistral Medium 3 blog table: 93.7.
mistral-medium-3,chartqa,82.6,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",True,official_blog,verified,Mistral Medium 3 blog table: 82.6.
llama-4-maverick,livecodebench_v6,28.7,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 28.7. Third-party self-test by Mistral (their internal eval pipeline).
llama-4-maverick,multipl_e_avg,76.4,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 76.4. Third-party self-test by Mistral (their internal eval pipeline).
llama-4-maverick,ruler_32k,94.8,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 94.8. Third-party self-test by Mistral (their internal eval pipeline).
llama-4-maverick,ruler_128k,86.7,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 86.7. Third-party self-test by Mistral (their internal eval pipeline).
llama-4-maverick,docvqa,94.1,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 94.1. Third-party self-test by Mistral (their internal eval pipeline).
llama-4-maverick,ai2d,84.4,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 84.4. Third-party self-test by Mistral (their internal eval pipeline).
llama-4-maverick,chartqa,90.4,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 90.4. Third-party self-test by Mistral (their internal eval pipeline).
gpt-4o,humaneval,91.5,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 91.5. Third-party self-test by Mistral (their internal eval pipeline).
gpt-4o,livecodebench_v6,31.4,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 31.4. Third-party self-test by Mistral (their internal eval pipeline).
gpt-4o,multipl_e_avg,79.8,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 79.8. Third-party self-test by Mistral (their internal eval pipeline).
gpt-4o,arena_hard,95.4,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 95.4. Third-party self-test by Mistral (their internal eval pipeline).
gpt-4o,math_500,76.4,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 76.4. Third-party self-test by Mistral (their internal eval pipeline).
gpt-4o,mmlu_pro,75.8,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""5-shot CoT"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 75.8. Third-party self-test by Mistral (their internal eval pipeline).
gpt-4o,ruler_32k,96.0,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 96.0. Third-party self-test by Mistral (their internal eval pipeline).
gpt-4o,ruler_128k,88.9,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 88.9. Third-party self-test by Mistral (their internal eval pipeline).
gpt-4o,docvqa,85.9,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 85.9. Third-party self-test by Mistral (their internal eval pipeline).
gpt-4o,ai2d,93.3,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 93.3. Third-party self-test by Mistral (their internal eval pipeline).
gpt-4o,chartqa,86.0,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 86.0. Third-party self-test by Mistral (their internal eval pipeline).
claude-sonnet-3.7,humaneval,92.1,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 92.1. Third-party self-test by Mistral (their internal eval pipeline).
claude-sonnet-3.7,livecodebench_v6,36.0,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 36.0. Third-party self-test by Mistral (their internal eval pipeline).
claude-sonnet-3.7,multipl_e_avg,83.4,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 83.4. Third-party self-test by Mistral (their internal eval pipeline).
claude-sonnet-3.7,arena_hard,93.2,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 93.2. Third-party self-test by Mistral (their internal eval pipeline).
claude-sonnet-3.7,ifeval,91.8,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 91.8. Third-party self-test by Mistral (their internal eval pipeline).
claude-sonnet-3.7,math_500,83.0,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 83.0. Third-party self-test by Mistral (their internal eval pipeline).
claude-sonnet-3.7,gpqa_diamond,69.7,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""5-shot CoT"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 69.7. Third-party self-test by Mistral (their internal eval pipeline).
claude-sonnet-3.7,mmlu_pro,80.0,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""5-shot CoT"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 80.0. Third-party self-test by Mistral (their internal eval pipeline).
claude-sonnet-3.7,ruler_32k,95.7,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 95.7. Third-party self-test by Mistral (their internal eval pipeline).
claude-sonnet-3.7,ruler_128k,93.8,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 93.8. Third-party self-test by Mistral (their internal eval pipeline).
claude-sonnet-3.7,mmmu,71.3,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 71.3. Third-party self-test by Mistral (their internal eval pipeline).
claude-sonnet-3.7,docvqa,84.3,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 84.3. Third-party self-test by Mistral (their internal eval pipeline).
claude-sonnet-3.7,ai2d,78.8,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 78.8. Third-party self-test by Mistral (their internal eval pipeline).
claude-sonnet-3.7,chartqa,76.3,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 76.3. Third-party self-test by Mistral (their internal eval pipeline).
command-a,livecodebench_v6,26.3,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 26.3. Third-party self-test by Mistral (their internal eval pipeline).
command-a,multipl_e_avg,73.1,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 73.1. Third-party self-test by Mistral (their internal eval pipeline).
command-a,ruler_32k,95.6,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 95.6. Third-party self-test by Mistral (their internal eval pipeline).
command-a,ruler_128k,91.2,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 91.2. Third-party self-test by Mistral (their internal eval pipeline).
deepseek-3.1,humaneval,93.3,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 93.3. Third-party self-test by Mistral (their internal eval pipeline).
deepseek-3.1,livecodebench_v6,42.9,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 42.9. Third-party self-test by Mistral (their internal eval pipeline).
deepseek-3.1,multipl_e_avg,84.9,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 84.9. Third-party self-test by Mistral (their internal eval pipeline).
deepseek-3.1,arena_hard,97.3,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 97.3. Third-party self-test by Mistral (their internal eval pipeline).
deepseek-3.1,ifeval,89.1,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 89.1. Third-party self-test by Mistral (their internal eval pipeline).
deepseek-3.1,math_500,93.8,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 93.8. Third-party self-test by Mistral (their internal eval pipeline).
deepseek-3.1,gpqa_diamond,61.1,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""5-shot CoT"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 61.1. Third-party self-test by Mistral (their internal eval pipeline).
deepseek-3.1,mmlu_pro,81.1,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""5-shot CoT"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 81.1. Third-party self-test by Mistral (their internal eval pipeline).
deepseek-3.1,ruler_32k,95.8,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 95.8. Third-party self-test by Mistral (their internal eval pipeline).
deepseek-3.1,ruler_128k,91.9,https://mistral.ai/news/mistral-medium-3,"{""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",False,official_blog,verified,Mistral Medium 3 blog table: 91.9. Third-party self-test by Mistral (their internal eval pipeline).
command-r7b,mmlu,65.2,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (command-r7b column): mmlu=65.2. Third-party Cohere reproduction, mc=false."
command-a,math,80.0,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Command A paper (command-a column): math=80.0.
gpt-4o-0513,math,70.2,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (gpt-4o-0513 column): math=70.2. Third-party Cohere reproduction, mc=false."
deepseek-v3,math,68.5,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (deepseek-v3 column): math=68.5. Third-party Cohere reproduction, mc=false."
command-r7b,math,59.1,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (command-r7b column): math=59.1. Third-party Cohere reproduction, mc=false."
command-r7b,ifeval,77.9,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (command-r7b column): ifeval=77.9. Third-party Cohere reproduction, mc=false."
command-a,gpqa_main,50.8,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Command A paper (command-a column): gpqa=50.8.
gpt-4o-0513,gpqa_main,59.1,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (gpt-4o-0513 column): gpqa=59.1. Third-party Cohere reproduction, mc=false."
deepseek-v3,gpqa_main,53.6,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (deepseek-v3 column): gpqa=53.6. Third-party Cohere reproduction, mc=false."
llama-3.3-70b-instruct,gpqa_main,50.5,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (llama-3.3-70b-instruct column): gpqa=50.5. Third-party Cohere reproduction, mc=false."
command-r7b,gpqa_main,26.3,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (command-r7b column): gpqa=26.3. Third-party Cohere reproduction, mc=false."
llama-3.1-8b,gpqa_main,23.4,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (llama-3.1-8b column): gpqa=23.4. Third-party Cohere reproduction, mc=false."
ministral-8b,gpqa_main,23.4,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (ministral-8b column): gpqa=23.4. Third-party Cohere reproduction, mc=false."
command-a,mbpp_plus,86.2,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Command A paper (command-a column): mbpp_plus=86.2.
gpt-4o-0513,mbpp_plus,89.9,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (gpt-4o-0513 column): mbpp_plus=89.9. Third-party Cohere reproduction, mc=false."
deepseek-v3,mbpp_plus,86.5,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (deepseek-v3 column): mbpp_plus=86.5. Third-party Cohere reproduction, mc=false."
llama-3.3-70b-instruct,mbpp_plus,84.4,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (llama-3.3-70b-instruct column): mbpp_plus=84.4. Third-party Cohere reproduction, mc=false."
command-r7b,mbpp_plus,72.0,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (command-r7b column): mbpp_plus=72.0. Third-party Cohere reproduction, mc=false."
llama-3.1-8b,mbpp_plus,72.8,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (llama-3.1-8b column): mbpp_plus=72.8. Third-party Cohere reproduction, mc=false."
ministral-8b,mbpp_plus,61.1,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (ministral-8b column): mbpp_plus=61.1. Third-party Cohere reproduction, mc=false."
command-a,bird_sql,59.5,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Command A paper (command-a column): bird_sql=59.5.
gpt-4o-0513,bird_sql,53.1,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (gpt-4o-0513 column): bird_sql=53.1. Third-party Cohere reproduction, mc=false."
deepseek-v3,bird_sql,50.5,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (deepseek-v3 column): bird_sql=50.5. Third-party Cohere reproduction, mc=false."
llama-3.3-70b-instruct,bird_sql,58.0,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (llama-3.3-70b-instruct column): bird_sql=58.0. Third-party Cohere reproduction, mc=false."
command-r7b,bird_sql,42.2,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (command-r7b column): bird_sql=42.2. Third-party Cohere reproduction, mc=false."
llama-3.1-8b,bird_sql,41.9,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (llama-3.1-8b column): bird_sql=41.9. Third-party Cohere reproduction, mc=false."
ministral-8b,bird_sql,33.2,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (ministral-8b column): bird_sql=33.2. Third-party Cohere reproduction, mc=false."
command-a,repoqa,92.6,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Command A paper (command-a column): repoqa=92.6.
gpt-4o-0513,repoqa,92.2,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (gpt-4o-0513 column): repoqa=92.2. Third-party Cohere reproduction, mc=false."
deepseek-v3,repoqa,91.2,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (deepseek-v3 column): repoqa=91.2. Third-party Cohere reproduction, mc=false."
llama-3.3-70b-instruct,repoqa,85.6,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (llama-3.3-70b-instruct column): repoqa=85.6. Third-party Cohere reproduction, mc=false."
command-r7b,repoqa,69.6,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (command-r7b column): repoqa=69.6. Third-party Cohere reproduction, mc=false."
llama-3.1-8b,repoqa,73.6,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (llama-3.1-8b column): repoqa=73.6. Third-party Cohere reproduction, mc=false."
ministral-8b,repoqa,62.0,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (ministral-8b column): repoqa=62.0. Third-party Cohere reproduction, mc=false."
command-a,ntrex,68.8,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Command A paper (command-a column): ntrex=68.8.
gpt-4o-0513,ntrex,69.8,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (gpt-4o-0513 column): ntrex=69.8. Third-party Cohere reproduction, mc=false."
deepseek-v3,ntrex,71.0,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (deepseek-v3 column): ntrex=71.0. Third-party Cohere reproduction, mc=false."
llama-3.3-70b-instruct,ntrex,62.5,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (llama-3.3-70b-instruct column): ntrex=62.5. Third-party Cohere reproduction, mc=false."
command-r7b,ntrex,48.1,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (command-r7b column): ntrex=48.1. Third-party Cohere reproduction, mc=false."
llama-3.1-8b,ntrex,49.2,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (llama-3.1-8b column): ntrex=49.2. Third-party Cohere reproduction, mc=false."
ministral-8b,ntrex,36.8,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (ministral-8b column): ntrex=36.8. Third-party Cohere reproduction, mc=false."
command-a,bfcl,63.8,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Command A paper (command-a column): bfcl=63.8.
gpt-4o-0513,bfcl,58.6,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (gpt-4o-0513 column): bfcl=58.6. Third-party Cohere reproduction, mc=false."
deepseek-v3,bfcl,72.1,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (deepseek-v3 column): bfcl=72.1. Third-party Cohere reproduction, mc=false."
llama-3.3-70b-instruct,bfcl,51.4,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (llama-3.3-70b-instruct column): bfcl=51.4. Third-party Cohere reproduction, mc=false."
command-r7b,bfcl,52.2,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (command-r7b column): bfcl=52.2. Third-party Cohere reproduction, mc=false."
llama-3.1-8b,bfcl,50.9,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (llama-3.1-8b column): bfcl=50.9. Third-party Cohere reproduction, mc=false."
ministral-8b,bfcl,51.8,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (ministral-8b column): bfcl=51.8. Third-party Cohere reproduction, mc=false."
command-a,tau1_bench_avg,51.7,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Command A paper (command-a column): taubench_avg=51.7.
gpt-4o-0513,tau1_bench_avg,39.1,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (gpt-4o-0513 column): taubench_avg=39.1. Third-party Cohere reproduction, mc=false."
deepseek-v3,tau1_bench_avg,51.2,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (deepseek-v3 column): taubench_avg=51.2. Third-party Cohere reproduction, mc=false."
llama-3.3-70b-instruct,tau1_bench_avg,21.0,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (llama-3.3-70b-instruct column): taubench_avg=21.0. Third-party Cohere reproduction, mc=false."
mistral-large-2,mmlu,85.2,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (mistral-large-2 column): mmlu=85.2. Third-party Cohere reproduction, mc=false."
gemini-2.0-pro,mmlu,89.3,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (gemini-2.0-pro column): mmlu=89.3. Third-party Cohere reproduction, mc=false."
gemini-1.5-flash-8b,mmlu,74.8,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (gemini-1.5-flash-8b column): mmlu=74.8. Third-party Cohere reproduction, mc=false."
mistral-large-2,mmlu_pro,67.9,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (mistral-large-2 column): mmlu_pro=67.9. Third-party Cohere reproduction, mc=false."
command-r7b,mmlu_pro,42.4,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (command-r7b column): mmlu_pro=42.4. Third-party Cohere reproduction, mc=false."
llama-3.1-8b,mmlu_pro,46.5,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (llama-3.1-8b column): mmlu_pro=46.5. Third-party Cohere reproduction, mc=false."
ministral-8b,mmlu_pro,43.0,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (ministral-8b column): mmlu_pro=43.0. Third-party Cohere reproduction, mc=false."
gemini-1.5-flash-8b,mmlu_pro,48.4,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (gemini-1.5-flash-8b column): mmlu_pro=48.4. Third-party Cohere reproduction, mc=false."
llama-3.1-405b-instruct,gpqa_main,49.0,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (llama-3.1-405b-instruct column): gpqa=49.0. Third-party Cohere reproduction, mc=false."
mistral-large-2,gpqa_main,48.6,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (mistral-large-2 column): gpqa=48.6. Third-party Cohere reproduction, mc=false."
claude-3.5-sonnet,gpqa_main,65.0,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (claude-3.5-sonnet column): gpqa=65.0. Third-party Cohere reproduction, mc=false."
gemini-2.0-pro,gpqa_main,64.7,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (gemini-2.0-pro column): gpqa=64.7. Third-party Cohere reproduction, mc=false."
gemma-2-9b,gpqa_main,31.3,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (gemma-2-9b column): gpqa=31.3. Third-party Cohere reproduction, mc=false."
gemini-1.5-flash-8b,gpqa_main,31.6,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (gemini-1.5-flash-8b column): gpqa=31.6. Third-party Cohere reproduction, mc=false."
mistral-large-2,ifeval,83.8,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (mistral-large-2 column): ifeval=83.8. Third-party Cohere reproduction, mc=false."
gemini-2.0-pro,ifeval,87.3,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (gemini-2.0-pro column): ifeval=87.3. Third-party Cohere reproduction, mc=false."
gemini-1.5-flash-8b,ifeval,88.0,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (gemini-1.5-flash-8b column): ifeval=88.0. Third-party Cohere reproduction, mc=false."
command-a,infobench,94.9,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Command A paper (command-a column): infobench=94.9.
gpt-4o-0513,infobench,94.0,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (gpt-4o-0513 column): infobench=94.0. Third-party Cohere reproduction, mc=false."
deepseek-v3,infobench,94.3,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (deepseek-v3 column): infobench=94.3. Third-party Cohere reproduction, mc=false."
llama-3.3-70b-instruct,infobench,92.8,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (llama-3.3-70b-instruct column): infobench=92.8. Third-party Cohere reproduction, mc=false."
llama-3.1-405b-instruct,infobench,93.9,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (llama-3.1-405b-instruct column): infobench=93.9. Third-party Cohere reproduction, mc=false."
mistral-large-2,infobench,93.3,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (mistral-large-2 column): infobench=93.3. Third-party Cohere reproduction, mc=false."
claude-3.5-sonnet,infobench,93.9,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (claude-3.5-sonnet column): infobench=93.9. Third-party Cohere reproduction, mc=false."
gemini-2.0-pro,infobench,92.2,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (gemini-2.0-pro column): infobench=92.2. Third-party Cohere reproduction, mc=false."
command-r7b,infobench,85.6,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (command-r7b column): infobench=85.6. Third-party Cohere reproduction, mc=false."
llama-3.1-8b,infobench,90.1,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (llama-3.1-8b column): infobench=90.1. Third-party Cohere reproduction, mc=false."
ministral-8b,infobench,88.3,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (ministral-8b column): infobench=88.3. Third-party Cohere reproduction, mc=false."
gemma-2-9b,infobench,87.2,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (gemma-2-9b column): infobench=87.2. Third-party Cohere reproduction, mc=false."
gemini-1.5-flash-8b,infobench,88.3,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (gemini-1.5-flash-8b column): infobench=88.3. Third-party Cohere reproduction, mc=false."
mistral-large-2,mbpp_plus,84.7,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (mistral-large-2 column): mbpp_plus=84.7. Third-party Cohere reproduction, mc=false."
qwen2.5-72b-instruct,mbpp_plus,88.6,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (qwen2.5-72b-instruct column): mbpp_plus=88.6. Third-party Cohere reproduction, mc=false."
llama-3.1-405b-instruct,mbpp_plus,88.6,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (llama-3.1-405b-instruct column): mbpp_plus=88.6. Third-party Cohere reproduction, mc=false."
command-r7b,livecodebench,9.0,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (command-r7b column): livecodebench=9.0. Third-party Cohere reproduction, mc=false."
llama-3.3-70b-instruct,livecodebench,32.9,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (llama-3.3-70b-instruct column): livecodebench=32.9. Third-party Cohere reproduction, mc=false."
mistral-large-2,livecodebench,26.7,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (mistral-large-2 column): livecodebench=26.7. Third-party Cohere reproduction, mc=false."
command-r7b,bigcodebench,30.9,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (command-r7b column): bigcodebench=30.9. Third-party Cohere reproduction, mc=false."
llama-3.3-70b-instruct,bigcodebench,46.9,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (llama-3.3-70b-instruct column): bigcodebench=46.9. Third-party Cohere reproduction, mc=false."
mistral-large-2,bigcodebench,44.7,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (mistral-large-2 column): bigcodebench=44.7. Third-party Cohere reproduction, mc=false."
qwen2.5-72b-instruct,bigcodebench,45.8,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (qwen2.5-72b-instruct column): bigcodebench=45.8. Third-party Cohere reproduction, mc=false."
llama-3.1-405b-instruct,bigcodebench,46.2,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (llama-3.1-405b-instruct column): bigcodebench=46.2. Third-party Cohere reproduction, mc=false."
command-r7b,humaneval,50.7,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (command-r7b column): humaneval=50.7. Third-party Cohere reproduction, mc=false."
mistral-large-2,humaneval,82.9,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (mistral-large-2 column): humaneval=82.9. Third-party Cohere reproduction, mc=false."
mistral-large-2,repoqa,88.0,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (mistral-large-2 column): repoqa=88.0. Third-party Cohere reproduction, mc=false."
qwen2.5-72b-instruct,repoqa,83.2,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (qwen2.5-72b-instruct column): repoqa=83.2. Third-party Cohere reproduction, mc=false."
llama-3.1-405b-instruct,repoqa,90.4,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (llama-3.1-405b-instruct column): repoqa=90.4. Third-party Cohere reproduction, mc=false."
command-r7b,swe_bench_verified,3.6,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (command-r7b column): swe_bench_verified=3.6. Third-party Cohere reproduction, mc=false."
llama-3.3-70b-instruct,swe_bench_verified,29.4,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (llama-3.3-70b-instruct column): swe_bench_verified=29.4. Third-party Cohere reproduction, mc=false."
mistral-large-2,swe_bench_verified,30.0,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (mistral-large-2 column): swe_bench_verified=30.0. Third-party Cohere reproduction, mc=false."
command-a,aider_polyglot,14.7,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Command A paper (command-a column): aider_polyglot=14.7.
command-r7b,aider_polyglot,2.7,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (command-r7b column): aider_polyglot=2.7. Third-party Cohere reproduction, mc=false."
llama-3.3-70b-instruct,aider_polyglot,8.4,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (llama-3.3-70b-instruct column): aider_polyglot=8.4. Third-party Cohere reproduction, mc=false."
mistral-large-2,aider_polyglot,16.0,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (mistral-large-2 column): aider_polyglot=16.0. Third-party Cohere reproduction, mc=false."
qwen2.5-72b-instruct,aider_polyglot,8.0,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (qwen2.5-72b-instruct column): aider_polyglot=8.0. Third-party Cohere reproduction, mc=false."
llama-3.1-405b-instruct,aider_polyglot,13.8,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (llama-3.1-405b-instruct column): aider_polyglot=13.8. Third-party Cohere reproduction, mc=false."
deepseek-v3,aider_polyglot,49.6,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (deepseek-v3 column): aider_polyglot=49.6. Third-party Cohere reproduction, mc=false."
mistral-large-2,bird_sql,50.0,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (mistral-large-2 column): bird_sql=50.0. Third-party Cohere reproduction, mc=false."
qwen2.5-72b-instruct,bird_sql,50.1,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (qwen2.5-72b-instruct column): bird_sql=50.1. Third-party Cohere reproduction, mc=false."
llama-3.1-405b-instruct,bird_sql,59.4,https://cohere.com/research/papers/command-a-technical-report.pdf,"{""effort"":""n/a"",""harness"":""Cohere internal eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,"Command A paper (llama-3.1-405b-instruct column): bird_sql=59.4. Third-party Cohere reproduction, mc=false."
internlm3-8b,cmmlu,83.1,https://github.com/InternLM/InternLM,"{""effort"":""default"",""harness"":""OpenCompass"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""tools"":""none""}",True,model_card,verified,CMMLU 0-shot = 83.1. No * (non-thinking). GitHub README table.
internlm3-8b,drop,83.1,https://github.com/InternLM/InternLM,"{""effort"":""default"",""harness"":""OpenCompass"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""tools"":""none""}",True,model_card,verified,DROP 0-shot = 83.1. No * (non-thinking). GitHub README table.
internlm3-8b,hellaswag,91.2,https://github.com/InternLM/InternLM,"{""effort"":""default"",""harness"":""OpenCompass"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""tools"":""none""}",True,model_card,verified,HellaSwag 10-shot = 91.2. No * (non-thinking). GitHub README table.
internlm3-8b,korbench,56.4,https://github.com/InternLM/InternLM,"{""effort"":""default"",""harness"":""OpenCompass"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""tools"":""none""}",True,model_card,verified,"KOR-Bench 0-shot = 56.4. No * (non-thinking). GitHub README table. Benchmark labeled KOR-Bench in source, maps to korbench in BP."
internlm3-8b,alpacaeval_2,51.1,https://github.com/InternLM/InternLM,"{""effort"":""default"",""harness"":""OpenCompass"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""tools"":""none""}",True,model_card,verified,AlpacaEval 2.0 LC WinRate = 51.1. No * (non-thinking). GitHub README table.
internlm3-8b,wildbench,33.1,https://github.com/InternLM/InternLM,"{""effort"":""default"",""harness"":""OpenCompass"",""judge"":""LLM-as-judge"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""tools"":""none""}",True,model_card,verified,WildBench Raw Score = 33.1. No * (non-thinking). GitHub README.
internlm3-8b,mt_bench_101,8.59,https://github.com/InternLM/InternLM,"{""effort"":""default"",""harness"":""OpenCompass"",""judge"":""LLM-as-judge"",""mode"":""non-thinking"",""prompt_style"":""0-shot"",""sampling"":""pass@1"",""tools"":""none""}",True,model_card,verified,MT-Bench-101 Score 1-10 = 8.59. No * (non-thinking). GitHub README.
exaone-4.0-32b,mmlu_redux,92.3,https://arxiv.org/html/2507.11407v1,"{""effort"":""default"",""harness"":""LG AI Research internal"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.6, top_p=0.95, presence_penalty=1.5"",""tools"":""none""}",True,tech_report,verified,Reasoning mode Table 3: MMLU-Redux=92.3. World Knowledge benchmark.
exaone-4.0-32b,livecodebench_v5,72.6,https://arxiv.org/html/2507.11407v1,"{""effort"":""default"",""harness"":""LG AI Research internal"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg@4"",""temperature"":""0.6, top_p=0.95, presence_penalty=1.5"",""tools"":""none""}",True,tech_report,verified,Reasoning mode Table 3: LiveCodeBench v5=72.6. (moved from generic livecodebench cell)
exaone-4.0-32b,livecodebench_v6,66.7,https://arxiv.org/html/2507.11407v1,"{""effort"":""default"",""harness"":""LG AI Research internal"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg@4"",""temperature"":""0.6, top_p=0.95, presence_penalty=1.5"",""tools"":""none""}",True,tech_report,verified,Reasoning mode Table 3: LiveCodeBench v6=66.7. Also Table 7 (64K budget confirms 66.7).
exaone-4.0-32b,multi_if,73.5,https://arxiv.org/html/2507.11407v1,"{""effort"":""default"",""harness"":""LG AI Research internal"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.6, top_p=0.95, presence_penalty=1.5"",""tools"":""none""}",True,tech_report,verified,Reasoning mode Table 3: Multi-IF (EN)=73.5. Instruction following.
exaone-4.0-32b,bfcl_v3,63.9,https://arxiv.org/html/2507.11407v1,"{""effort"":""default"",""harness"":""LG AI Research internal"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.6, top_p=0.95, presence_penalty=1.5"",""tools"":""none""}",True,tech_report,verified,Reasoning mode Table 3: BFCL-v3=63.9. Agentic Tool Use.
exaone-4.0-32b,tau_bench_airline,51.5,https://arxiv.org/html/2507.11407v1,"{""effort"":""default"",""harness"":""LG AI Research internal"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg@4"",""temperature"":""0.6, top_p=0.95, presence_penalty=1.5"",""tools"":""none""}",True,tech_report,verified,Reasoning mode Table 3: Tau-Bench (Airline)=51.5. Agentic Tool Use.
exaone-4.0-32b,tau_bench_retail,62.8,https://arxiv.org/html/2507.11407v1,"{""effort"":""default"",""harness"":""LG AI Research internal"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg@4"",""temperature"":""0.6, top_p=0.95, presence_penalty=1.5"",""tools"":""none""}",True,tech_report,verified,Reasoning mode Table 3: Tau-Bench (Retail)=62.8. Agentic Tool Use.
claude-3.5-sonnet,tau_bench_retail,71.5,https://www.anthropic.com/news/claude-3-7-sonnet,"{""effort"":""default"",""harness"":""official"",""judge"":""official harness"",""mode"":""non-thinking"",""prompt_style"":""default + planning tool addendum (Airline)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (tau-bench official)""}",True,official_blog,verified,Anthropic 3.7 Sonnet blog cross-model comparison table: Claude 3.5 Sonnet (new) TAU-bench Retail=71.5%
claude-3.5-sonnet,tau_bench_airline,48.8,https://www.anthropic.com/news/claude-3-7-sonnet,"{""effort"":""default"",""harness"":""official"",""judge"":""official harness"",""mode"":""non-thinking"",""prompt_style"":""default + planning tool addendum (Airline)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (tau-bench official)""}",True,official_blog,verified,Anthropic 3.7 Sonnet blog cross-model comparison table: Claude 3.5 Sonnet (new) TAU-bench Airline=48.8%
claude-3.5-sonnet,mmmlu,82.1,https://www.anthropic.com/news/claude-3-7-sonnet,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,Anthropic 3.7 Sonnet blog cross-model: Claude 3.5 Sonnet (new) MMMLU=82.1%
claude-3.5-sonnet,mmmu,70.4,https://www.anthropic.com/news/claude-3-7-sonnet,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""prompt_style"":""zeroshot-cot"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,Anthropic 3.7 Sonnet blog cross-model: Claude 3.5 Sonnet (new) MMMU=70.4%
kimi-k2.5,hle_tools,50.2,https://huggingface.co/moonshotai/Kimi-K2.5,"""thinking mode, tool-augmented, pass@1""",True,model_card,verified,"First-party, directly from Kimi K2.5 HF model card evaluation table."
gpt-5.2,hle_tools,45.5,https://huggingface.co/moonshotai/Kimi-K2.5,"""with tools, pass@1""",False,model_card,verified,"Reported in Kimi K2.5 HF model card (no asterisk = Moonshot used publicly available official score). Reference is Kimi card, not GPT primary source."
claude-opus-4.5,hle_tools,43.2,https://huggingface.co/moonshotai/Kimi-K2.5,"""with tools, pass@1""",False,model_card,verified,Reported in Kimi K2.5 HF model card (no asterisk = publicly available official score). Reference is Kimi card.
gemini-3-pro,hle_tools,45.8,https://huggingface.co/moonshotai/Kimi-K2.5,"""with tools, pass@1""",False,model_card,verified,Reported in Kimi K2.5 HF model card (no asterisk = publicly available official score). Reference is Kimi card.
deepseek-v3.2,hle_tools,40.8,https://huggingface.co/moonshotai/Kimi-K2.5,"""text-only subset, with tools, pass@1""",False,model_card,verified,†: DeepSeek V3.2 HLE (w/tools) score corresponds to text-only subset (vision not supported). Reported in Kimi K2.5 HF model card.
kimi-k2.5,zerobench_tools,11.0,https://huggingface.co/moonshotai/Kimi-K2.5,"""thinking mode, tool-augmented, pass@1""",True,model_card,verified,"First-party, directly from Kimi K2.5 HF model card."
gpt-5.2,zerobench_tools,7.0,https://huggingface.co/moonshotai/Kimi-K2.5,"""tool-augmented, pass@1; re-evaluated by Moonshot""",False,model_card,verified,*: Score re-evaluated by Moonshot AI under Kimi K2.5 evaluation conditions (no publicly available score).
claude-opus-4.5,zerobench_tools,9.0,https://huggingface.co/moonshotai/Kimi-K2.5,"""tool-augmented, pass@1; re-evaluated by Moonshot""",False,model_card,verified,*: Score re-evaluated by Moonshot AI under Kimi K2.5 evaluation conditions.
gemini-3-pro,zerobench_tools,12.0,https://huggingface.co/moonshotai/Kimi-K2.5,"""tool-augmented, pass@1; re-evaluated by Moonshot""",False,model_card,verified,*: Score re-evaluated by Moonshot AI under Kimi K2.5 evaluation conditions.
deepseek-v3.2,widesearch,32.5,https://huggingface.co/moonshotai/Kimi-K2.5,"""agentic search, pass@1; re-evaluated by Moonshot""",False,model_card,verified,*: Score re-evaluated by Moonshot AI under Kimi K2.5 evaluation conditions (WideSearch benchmark). DeepSeek's own report does not include this benchmark.
gpt-5,toolathlon,29.0,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""context"":""128K"",""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",False,tech_report,verified,DS V3.2 tech report Table 2: GPT-5-High=29.0. Third-party self-test by DeepSeek.
kimi-k2-thinking,toolathlon,17.6,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""context"":""128K"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",False,tech_report,verified,DS V3.2 tech report Table 2: Kimi-K2-Thinking=17.6. Third-party self-test by DeepSeek.
minimax-m2,toolathlon,16.0,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""context"":""128K"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",False,tech_report,verified,DS V3.2 tech report Table 2: MiniMax-M2=16.0. Third-party self-test by DeepSeek.
gemini-3-pro,tau2_bench_avg,85.4,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""context"":""128K"",""effort"":""high"",""harness"":""internal"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",False,tech_report,verified,"DS V3.2 tech report Table 2: Gemini-3.0-Pro=85.4. DS uses the test model itself as user agent (non-standard harness), not the official simulated-user harness. Third-party self-test."
kimi-k2-thinking,tau2_bench_avg,74.3,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""context"":""128K"",""effort"":""default"",""harness"":""internal"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",False,tech_report,verified,DS V3.2 tech report Table 2: Kimi-K2-Thinking=74.3. DS uses the test model itself as user agent (non-standard harness). Third-party self-test.
gpt-5,mcpmark,50.9,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""context"":""128K"",""effort"":""high"",""harness"":""internal"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",False,tech_report,verified,DS V3.2 tech report Table 2: GPT-5-High=50.9. DS uses internal MCP environment (not official setting). Third-party self-test.
kimi-k2-thinking,mcpmark,20.4,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""context"":""128K"",""effort"":""default"",""harness"":""internal"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",False,tech_report,verified,DS V3.2 tech report Table 2: Kimi-K2-Thinking=20.4. DS uses internal MCP environment. Third-party self-test.
minimax-m2,mcpmark,24.4,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""context"":""128K"",""effort"":""default"",""harness"":""internal"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",False,tech_report,verified,DS V3.2 tech report Table 2: MiniMax-M2=24.4. DS uses internal MCP environment. Third-party self-test.
deepseek-v3.2,mcpmark,38.0,https://huggingface.co/deepseek-ai/DeepSeek-V3.2/resolve/main/assets/paper.pdf,"{""context"":""128K"",""effort"":""default"",""harness"":""internal"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic""}",False,tech_report,verified,DS V3.2 tech report Table 2: DeepSeek-V3.2-Thinking=38.0. DS uses internal MCP environment (not official MCP-Mark setting).
gemini-3-flash,omnidocbench_1.5,0.121,,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: OmniDocBench 1.5 overall edit distance 0.121 (lower=better)
gemini-2.5-flash,omnidocbench_1.5,0.154,,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: OmniDocBench 1.5 overall edit distance 0.154
gemini-2.5-pro,omnidocbench_1.5,0.145,,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: OmniDocBench 1.5 overall edit distance 0.145
claude-sonnet-4.5,omnidocbench_1.5,0.145,,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: OmniDocBench 1.5 0.145 (third-party self-test by Google DeepMind)
gemini-3-flash,video_mmmu,86.9,,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: Video-MMMU (Knowledge acquisition from videos) 86.9%
gemini-3-pro,video_mmmu,87.6,,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,"DeepMind /models/gemini/flash/ Performance table: Video-MMMU 87.6% (cross-model comparison; Flash page column label ""Gemini 3 Pro Thinking"" does not specify ""High"" effort — matches_canonical=false)"
gemini-2.5-flash,video_mmmu,79.2,,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: Video-MMMU 79.2%
gemini-2.5-pro,video_mmmu,83.6,,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: Video-MMMU 83.6%
claude-sonnet-4.5,video_mmmu,77.8,,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: Video-MMMU 77.8% (third-party self-test by Google DeepMind)
gpt-5.2,video_mmmu,85.9,,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: Video-MMMU 85.9% (third-party self-test by Google DeepMind)
gemini-3-flash,tau2_bench_avg,90.2,,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: τ2-bench avg (agentic tool use) 90.2%
gemini-2.5-flash,tau2_bench_avg,79.5,,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,DeepMind /models/gemini/flash/ Performance table: τ2-bench avg 79.5%
grok-4.1,vending_bench_2,1107.0,,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"DeepMind /models/gemini/flash/ Performance table: Vending-Bench 2 net worth (mean) $1,107 (Grok 4.1 Fast Reasoning column)"
gpt-4o,arc_agi_1,4.5,https://arcprize.org/leaderboard,"{""effort"":""n/a"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""non-thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""Base LLM"",""tools"":""none""}",True,leaderboard,verified,arcprize.org leaderboard audit R5b: GPT-4o Base LLM on leaderboard
gpt-4o,arc_agi_2,0.0,https://arcprize.org/leaderboard,"{""effort"":""n/a"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""non-thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""Base LLM"",""tools"":""none""}",True,leaderboard,verified,arcprize.org leaderboard audit R5b: GPT-4o Base LLM on leaderboard
gpt-4o-mini,arc_agi_2,0.0,https://arcprize.org/leaderboard,"{""effort"":""n/a"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""non-thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""Base LLM"",""tools"":""none""}",True,leaderboard,verified,arcprize.org leaderboard audit R5b: GPT-4o-mini Base LLM on leaderboard (arc_agi_1 N/A)
o1-mini,arc_agi_1,14.0,https://arcprize.org/leaderboard,"{""effort"":""default"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""CoT"",""tools"":""none""}",False,leaderboard,verified,arcprize.org leaderboard audit R5b: o1-mini on leaderboard
o1-mini,arc_agi_2,0.8,https://arcprize.org/leaderboard,"{""effort"":""default"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""CoT"",""tools"":""none""}",False,leaderboard,verified,arcprize.org leaderboard audit R5b: o1-mini on leaderboard
gemini-1.5-pro,arc_agi_2,0.8,https://arcprize.org/leaderboard,"{""effort"":""n/a"",""harness"":""official (arcprize.org)"",""judge"":""rule-based"",""mode"":""non-thinking"",""sampling"":""pass@1 (single-pass)"",""system_type"":""Base LLM"",""tools"":""none""}",True,leaderboard,verified,arcprize.org leaderboard audit R5b: Gemini 1.5 Pro on leaderboard (arc_agi_1 N/A)
grok-4.20,terminal_bench,57.3,https://www.tbench.ai/leaderboard/terminal-bench/2.0,{},,leaderboard,verified,"grok-cli third-party agent, score=57.3% ± N/A. Only available entry on tbench 2.0."
minimax-m2.1,terminal_bench,29.2,https://www.tbench.ai/leaderboard/terminal-bench/2.0,{},,leaderboard,verified,"Confirmed 29.2% on tbench.ai Terminal-Bench 2.0 (Terminus 2 scaffold, 2025-12-23)."
minimax-m2.5,terminal_bench,42.2,https://www.tbench.ai/leaderboard/terminal-bench/2.0,{},,leaderboard,verified,"Confirmed 42.2% on tbench.ai Terminal-Bench 2.0 (Terminus 2 scaffold, 2026-02-23). minimax-m2.5 listed as ""Minimax m2.5""."
minimax-m2,hle_tools,31.8,https://huggingface.co/MiniMaxAI/MiniMax-M2,"{""effort"":""default"",""harness"":""WebExplorer agent framework"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""search + python (text-only HLE subset)""}",True,model_card,verified,"MiniMax M2 model card Table 1: HLE (w/ tools)=31.8. Uses search tools + Python tool (Jupyter), text-only HLE subset."
minimax-m2.7,multi_swe_bench,52.7,https://huggingface.co/MiniMaxAI/MiniMax-M2.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,model_card,verified,MiniMax M2.7 model card: Multi-SWE-Bench=52.7.
minimax-m2.7,swe_bench_multilingual,76.5,https://huggingface.co/MiniMaxAI/MiniMax-M2.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic (bash + edit)""}",True,model_card,verified,MiniMax M2.7 model card: SWE-bench Multilingual=76.5.
minimax-m2.7,terminal_bench,57.0,https://huggingface.co/MiniMaxAI/MiniMax-M2.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,model_card,verified,MiniMax M2.7 model card: Terminal-Bench 2=57.0.
glm-4.7,hle_tools,42.8,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""search+code+web""}",True,official_blog,verified,GLM-4.7 blog (first-party): GLM-4.7=42.8. Also confirmed in z.ai/blog/glm-5 (third-party: 42.8).
glm-4.6,hle_tools,30.4,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""search+code+web""}",True,official_blog,verified,"GLM-4.7 blog (same provider Zhipu, third-party for GLM-4.6): GLM-4.6=30.4 in comparison table."
glm-5,hle_tools,50.4,https://z.ai/blog/glm-5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""search+code+web""}",True,official_blog,verified,GLM-5 blog (first-party): GLM-5=50.4. Also confirmed in z.ai/blog/glm-5.1 (50.4).
glm-4.7,tau2_bench_avg,87.4,https://z.ai/blog/glm-4.7,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""web+API""}",True,official_blog,verified,GLM-4.7 blog (first-party): τ²-Bench=87.4. Confirmed by z.ai/blog/glm-5 (third-party: 87.4).
glm-5,tau2_bench_avg,89.7,https://z.ai/blog/glm-5,"{""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""web+API""}",True,official_blog,verified,"GLM-5 blog (first-party): τ²-Bench=89.7. Confirmed by z.ai/blog/glm-5.1 (same value 89.7, via comparison)."
glm-4.6,browsecomp_cm,57.5,https://z.ai/blog/glm-4.7,"{""context_management"":""discard-all"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-4.7 blog (same provider Zhipu): GLM-4.6=57.5 in BrowseComp (w/ Context Manage) table.
glm-4.7,browsecomp_cm,67.5,https://z.ai/blog/glm-4.7,"{""context_management"":""discard-all"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-4.7 blog (first-party): GLM-4.7=67.5. Confirmed in z.ai/blog/glm-5 (third-party: 67.5).
glm-5,browsecomp_cm,75.9,https://z.ai/blog/glm-5,"{""context_management"":""discard-all"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,GLM-5 blog (first-party): GLM-5=75.9. Confirmed in z.ai/blog/glm-5.1 (75.9).
glm-5.1,browsecomp_cm,79.3,https://z.ai/blog/glm-5.1,"{""context_management"":""discard-all"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic""}",True,official_blog,verified,"GLM-5.1 blog (first-party): GLM-5.1=79.3. NOTE: BP browsecomp primary=79.3 (DeepSeek V4 Pro card) may be mislabeled as plain BrowseComp; GLM-5.1 blog gives plain=68, CM=79.3."
claude-opus-4,cybench,56.4,https://www-cdn.anthropic.com/07b2a3f9902ee19fe39a36ca638e5ae987bc64dd.pdf,"{""effort"":""default"",""harness"":""Anthropic internal (Kali-based VM)"",""judge"":""rule-based (CTF flag check)"",""mode"":""thinking"",""notes"":""Claude 4 System Card \u00a77.4.7: claude-opus-4 scored 22/39 on Cybench (1 challenge skipped due to infra/timing; benchmark has 40 total). Thinking mode not explicitly specified but consistent with canonical."",""prompt_style"":""default"",""sampling"":""pass@30"",""temperature"":""default"",""tools"":""agentic (Kali environment with code editor + Terminal Tool; pwntools/metasploit/ghidra)""}",True,model_card,verified,"From Claude 4 System Card (May 2025), §7.4.7 Cybench: 22/39 = 56.4%."
claude-sonnet-4,cybench,56.4,https://www-cdn.anthropic.com/07b2a3f9902ee19fe39a36ca638e5ae987bc64dd.pdf,"{""effort"":""default"",""harness"":""Anthropic internal (Kali-based VM)"",""judge"":""rule-based (CTF flag check)"",""mode"":""thinking"",""notes"":""Claude 4 System Card \u00a77.4.7: claude-sonnet-4 scored 22/39 on Cybench (1 challenge skipped due to infra/timing; benchmark has 40 total). Thinking mode not explicitly specified but consistent with canonical."",""prompt_style"":""default"",""sampling"":""pass@30"",""temperature"":""default"",""tools"":""agentic (Kali environment with code editor + Terminal Tool; pwntools/metasploit/ghidra)""}",True,model_card,verified,"From Claude 4 System Card (May 2025), §7.4.7 Cybench: 22/39 = 56.4%."
gpt-5-mini,aime_2025,91.1,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"AIME 2025 no tools, per GPT-5 dev blog table 0"
gpt-5-mini,frontiermath,22.1,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""python""}",True,official_blog,verified,"FrontierMath with Python tool, per GPT-5 dev blog table 0"
gpt-5-mini,gpqa_diamond,82.3,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"GPQA Diamond no tools, per GPT-5 dev blog table 0"
gpt-5-mini,hle,16.7,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"HLE no tools, per GPT-5 dev blog table 0"
gpt-5-mini,hmmt_feb_2025,87.8,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"HMMT 2025 no tools, per GPT-5 dev blog table 0"
gpt-5-mini,mmmu,81.6,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"MMMU, per GPT-5 dev blog table 1"
gpt-5-mini,mmmu_pro,74.1,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"MMMU-Pro avg, per GPT-5 dev blog table 1"
gpt-5-mini,charxiv_reasoning,75.5,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""python""}",True,official_blog,verified,"CharXiv reasoning python-enabled, per GPT-5 dev blog table 1"
gpt-5-mini,video_mmmu,82.5,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"VideoMMMU max-frame 256, per GPT-5 dev blog table 1"
gpt-5-mini,erqa,62.9,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"ERQA, per GPT-5 dev blog table 1"
gpt-5-mini,swelancer_freelance_dollars,75000.0,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"SWE-Lancer IC SWE Diamond Freelance $75K, per GPT-5 dev blog table 2"
gpt-5-mini,swe_bench_verified,71.0,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"SWE-bench Verified, per GPT-5 dev blog table 2"
gpt-5-mini,aider_polyglot_diff,71.6,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"Aider polyglot diff, per GPT-5 dev blog table 2"
gpt-5-mini,multichallenge_o3mini_grader,62.3,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"Scale Multichallenge o3-mini grader, per GPT-5 dev blog table 3"
gpt-5-mini,internal_api_if_hard,65.8,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"Internal API IF hard, per GPT-5 dev blog table 3"
gpt-5-mini,collie,98.5,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"COLLIE, per GPT-5 dev blog table 3"
gpt-5-mini,tau2_bench_airline,60.0,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"Tau2-bench airline, per GPT-5 dev blog table 4"
gpt-5-mini,tau2_bench_retail,78.3,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"Tau2-bench retail, per GPT-5 dev blog table 4"
gpt-5-mini,tau2_bench_telecom,74.1,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"Tau2-bench telecom, per GPT-5 dev blog table 4"
gpt-5-mini,mrcr_v2_2needle_128k,84.3,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"OpenAI-MRCR 2-needle 128k, per GPT-5 dev blog table 5"
gpt-5-mini,mrcr_v2_2needle_256k,58.8,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"OpenAI-MRCR 2-needle 256k, per GPT-5 dev blog table 5"
gpt-5-mini,graphwalks_bfs_0k_128k,73.4,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"Graphwalks BFS <128k, per GPT-5 dev blog table 5"
gpt-5-mini,graphwalks_parents_0k_128k,64.3,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"Graphwalks parents <128k, per GPT-5 dev blog table 5"
gpt-5-mini,browsecomp_long_context_128k,89.4,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"BrowseComp Long Context 128k, per GPT-5 dev blog table 5"
gpt-5-mini,browsecomp_long_context_256k,86.0,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"BrowseComp Long Context 256k, per GPT-5 dev blog table 5"
gpt-5-mini,video_mme,78.5,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"VideoMME long with subtitle, per GPT-5 dev blog table 5"
gpt-5-mini,longfact_concepts,0.7,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"LongFact-Concepts hallucination rate, per GPT-5 dev blog table 6"
gpt-5-mini,longfact_objects,1.3,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"LongFact-Objects hallucination rate, per GPT-5 dev blog table 6"
gpt-5-mini,factscore,3.5,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"FActScore hallucination rate, per GPT-5 dev blog table 6"
gpt-5-nano,aime_2025,85.2,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"AIME 2025 no tools, per GPT-5 dev blog table 0"
gpt-5-nano,frontiermath,9.6,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""python""}",True,official_blog,verified,"FrontierMath with Python tool, per GPT-5 dev blog table 0"
gpt-5-nano,gpqa_diamond,71.2,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"GPQA Diamond no tools, per GPT-5 dev blog table 0"
gpt-5-nano,hle,8.7,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"HLE no tools, per GPT-5 dev blog table 0"
gpt-5-nano,hmmt_feb_2025,75.6,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"HMMT 2025 no tools, per GPT-5 dev blog table 0"
gpt-5-nano,mmmu,75.6,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"MMMU, per GPT-5 dev blog table 1"
gpt-5-nano,mmmu_pro,62.6,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"MMMU-Pro avg, per GPT-5 dev blog table 1"
gpt-5-nano,charxiv_reasoning,62.7,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""python""}",True,official_blog,verified,"CharXiv reasoning python-enabled, per GPT-5 dev blog table 1"
gpt-5-nano,video_mmmu,66.8,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"VideoMMMU max-frame 256, per GPT-5 dev blog table 1"
gpt-5-nano,erqa,50.1,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"ERQA, per GPT-5 dev blog table 1"
gpt-5-nano,swelancer_freelance_dollars,49000.0,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"SWE-Lancer IC SWE Diamond Freelance $49K, per GPT-5 dev blog table 2"
gpt-5-nano,swe_bench_verified,54.7,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"SWE-bench Verified, per GPT-5 dev blog table 2"
gpt-5-nano,aider_polyglot_diff,48.4,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"Aider polyglot diff, per GPT-5 dev blog table 2"
gpt-5-nano,multichallenge_o3mini_grader,54.9,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"Scale Multichallenge o3-mini grader, per GPT-5 dev blog table 3"
gpt-5-nano,internal_api_if_hard,56.1,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"Internal API IF hard, per GPT-5 dev blog table 3"
gpt-5-nano,collie,96.9,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"COLLIE, per GPT-5 dev blog table 3"
gpt-5-nano,tau2_bench_airline,41.0,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"Tau2-bench airline, per GPT-5 dev blog table 4"
gpt-5-nano,tau2_bench_retail,62.3,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"Tau2-bench retail, per GPT-5 dev blog table 4"
gpt-5-nano,tau2_bench_telecom,35.5,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"Tau2-bench telecom, per GPT-5 dev blog table 4"
gpt-5-nano,mrcr_v2_2needle_128k,43.2,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"OpenAI-MRCR 2-needle 128k, per GPT-5 dev blog table 5"
gpt-5-nano,mrcr_v2_2needle_256k,34.9,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"OpenAI-MRCR 2-needle 256k, per GPT-5 dev blog table 5"
gpt-5-nano,graphwalks_bfs_0k_128k,64.0,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"Graphwalks BFS <128k, per GPT-5 dev blog table 5"
gpt-5-nano,graphwalks_parents_0k_128k,43.8,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"Graphwalks parents <128k, per GPT-5 dev blog table 5"
gpt-5-nano,browsecomp_long_context_128k,80.4,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"BrowseComp Long Context 128k, per GPT-5 dev blog table 5"
gpt-5-nano,browsecomp_long_context_256k,68.4,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"BrowseComp Long Context 256k, per GPT-5 dev blog table 5"
gpt-5-nano,video_mme,65.7,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"VideoMME long with subtitle, per GPT-5 dev blog table 5"
gpt-5-nano,longfact_concepts,1.0,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"LongFact-Concepts hallucination rate, per GPT-5 dev blog table 6"
gpt-5-nano,longfact_objects,2.8,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"LongFact-Objects hallucination rate, per GPT-5 dev blog table 6"
gpt-5-nano,factscore,7.3,https://openai.com/index/introducing-gpt-5-for-developers/,"{""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none""}",True,official_blog,verified,"FActScore hallucination rate, per GPT-5 dev blog table 6"
gpt-5.2-codex,swe_bench_pro,56.4,https://openai.com/index/introducing-gpt-5-3-codex/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none (per-bench override)""}",True,official_blog,verified,"SWE-Bench Pro (Public), per GPT-5.3-Codex blog table 0"
gpt-5.2-codex,terminal_bench,64.0,https://openai.com/index/introducing-gpt-5-3-codex/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none (per-bench override)""}",True,official_blog,verified,"Terminal-Bench 2.0, per GPT-5.3-Codex blog table 0"
gpt-5.2-codex,osworld,38.2,https://openai.com/index/introducing-gpt-5-3-codex/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none (per-bench override)""}",True,official_blog,verified,"OSWorld-Verified, per GPT-5.3-Codex blog table 0"
gpt-5.2-codex,cybersecurity_ctf,67.4,https://openai.com/index/introducing-gpt-5-3-codex/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none (per-bench override)""}",True,official_blog,verified,"Cybersecurity CTF, per GPT-5.3-Codex blog table 0"
gpt-5.2-codex,swelancer,76.0,https://openai.com/index/introducing-gpt-5-3-codex/,"{""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""sampling"":""pass@1"",""tools"":""none (per-bench override)""}",True,official_blog,verified,"SWE-Lancer IC Diamond, per GPT-5.3-Codex blog table 0"
claude-opus-4.7,frontiermath_tier4,22.9,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""default"",""harness"":""OpenAI internal"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,third_party,verified,GPT-5.5 blog table: FrontierMath Tier 4=22.9%; third-party self-test by OpenAI
claude-opus-4.7,gdpval_oai_woe,80.3,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""default"",""harness"":""OpenAI internal"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,third_party,verified,GPT-5.5 blog table: GDPval(wins or ties)=80.3%; third-party self-test by OpenAI
claude-opus-4.7,officeqa_pro,43.6,https://openai.com/index/introducing-gpt-5-5/,"{""effort"":""default"",""harness"":""OpenAI internal"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,third_party,verified,GPT-5.5 blog table: OfficeQA Pro=43.6%; third-party self-test by OpenAI
muse-spark,charxiv_reasoning,86.4,https://ai.meta.com/blog/introducing-muse-spark-msl/,"{""effort"":""default"",""harness"":""internal (Meta MSL)"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (single-pass)"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"Meta MSL blog main table: CharXiv Reasoning 86.4, Thinking mode."
muse-spark,erqa,64.7,https://ai.meta.com/blog/introducing-muse-spark-msl/,"{""effort"":""default"",""harness"":""internal (Meta MSL)"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (single-pass)"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"Meta MSL blog main table: ERQA 64.7, Thinking mode. 925 questions with GPT-4.1-genai judge."
muse-spark,simplevqa,71.3,https://ai.meta.com/blog/introducing-muse-spark-msl/,"{""effort"":""default"",""harness"":""internal (Meta MSL)"",""judge"":""rule-based"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (single-pass)"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"Meta MSL blog main table: SimpleVQA 71.3, Thinking mode."
muse-spark,screenspot_pro,84.1,https://ai.meta.com/blog/introducing-muse-spark-msl/,"{""effort"":""default"",""harness"":""internal (Meta MSL)"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""with-Python variant per methodology; canonical BP is no-tools"",""prompt_style"":""default"",""sampling"":""pass@1 (single-pass)"",""temperature"":""default"",""tools"":""python (cropping tool)""}",False,official_blog,verified,Meta MSL blog: ScreenSpot Pro 84.1 with Python cropping tool. matches_canonical=false: canonical uses no tools.
muse-spark,zerobench_tools,33.0,https://ai.meta.com/blog/introducing-muse-spark-msl/,"{""effort"":""default"",""harness"":""internal (Meta MSL)"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""pass@5 per zerobench leaderboard convention; with Python"",""prompt_style"":""default"",""sampling"":""pass@5"",""temperature"":""default"",""tools"":""python""}",False,official_blog,verified,"Meta MSL blog: ZeroBench 33.0, pass@5 with Python. matches_canonical=false: canonical BP uses pass@1."
muse-spark,hle_tools,50.4,https://ai.meta.com/blog/introducing-muse-spark-msl/,"{""effort"":""default"",""harness"":""internal (Meta MSL)"",""judge"":""gpt-o3-mini (LLM-as-judge)"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (single-pass)"",""temperature"":""default"",""tools"":""bash + browser (blocklist: hf.co, kaggle, pwc, lmsys)""}",True,official_blog,verified,"Meta MSL blog: HLE (With Tools) 50.4, Thinking mode. Bash and browser tools enabled per methodology."
muse-spark,healthbench_hard,42.8,https://ai.meta.com/blog/introducing-muse-spark-msl/,"{""effort"":""default"",""harness"":""internal (Meta MSL)"",""judge"":""gpt-4.1-genai (LLM-as-judge)"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (single-pass)"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"Meta MSL blog: HealthBench Hard 42.8, Thinking mode. 1,000 prompts, GPT-4.1-genai judge."
muse-spark,medxpertqa_text,52.6,https://ai.meta.com/blog/introducing-muse-spark-msl/,"{""effort"":""default"",""harness"":""internal (Meta MSL)"",""judge"":""gpt-oss-120b"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (single-pass)"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"Meta MSL blog: MedXpertQA Text 52.6, Thinking mode. 2,450 prompts, 10 choices, gpt-oss-120b judge."
muse-spark,medxpertqa_mm,78.4,https://ai.meta.com/blog/introducing-muse-spark-msl/,"{""effort"":""default"",""harness"":""internal (Meta MSL)"",""judge"":""gpt-oss-120b"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (single-pass)"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"Meta MSL blog: MedXpertQA MM 78.4, Thinking mode. 2,000 multimodal medical questions."
muse-spark,deepsearchqa_f1,74.8,https://ai.meta.com/blog/introducing-muse-spark-msl/,"{""effort"":""default"",""harness"":""internal (Meta MSL)"",""judge"":""gpt-oss-120b (LLM-as-judge for F1)"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (single-pass)"",""temperature"":""default"",""tools"":""agentic (browser: search, open, find)""}",True,official_blog,verified,"Meta MSL blog: DeepSearchQA F1 74.8, Thinking mode. Agentic browser tools, F1 metric."
muse-spark,tau2_bench_telecom,91.5,https://ai.meta.com/blog/introducing-muse-spark-msl/,"{""effort"":""default"",""harness"":""Artificial Analysis \u03c4\u00b2-Bench Telecom Leaderboard"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Results from Artificial Analysis leaderboard per Meta methodology"",""prompt_style"":""default"",""sampling"":""pass@1 (single-pass)"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,Meta MSL blog: τ²-bench Telecom 91.5. Sourced from Artificial Analysis leaderboard.
o1-mini,humaneval,92.4,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'o1-mini', rank 7, accessed 2026-04-29). Setting unknown."
qwen-2.5-32b,humaneval,88.4,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Qwen2.5 32B Instruct', rank 21, accessed 2026-04-29). Setting unknown."
gemma-3-12b,humaneval,85.4,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Gemma 3 12B', rank 34, accessed 2026-04-29). Setting unknown."
qwen-2.5-7b,humaneval,84.8,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Qwen2.5 7B Instruct', rank 37, accessed 2026-04-29). Setting unknown."
gemini-1.5-pro,humaneval,84.1,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Gemini 1.5 Pro', rank 39, accessed 2026-04-29). Setting unknown."
llama-3.1-70b,humaneval,80.5,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Llama 3.1 70B Instruct', rank 45, accessed 2026-04-29). Setting unknown."
gemini-1.5-flash,humaneval,74.3,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Gemini 1.5 Flash', rank 51, accessed 2026-04-29). Setting unknown."
llama-3.1-8b,humaneval,72.6,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Llama 3.1 8B Instruct', rank 54, accessed 2026-04-29). Setting unknown."
gemma-3-4b,humaneval,71.3,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Gemma 3 4B', rank 56, accessed 2026-04-29). Setting unknown."
gpt-3.5-turbo,humaneval,68.0,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'GPT-3.5 Turbo', rank 58, accessed 2026-04-29). Setting unknown."
gemma-2-27b,humaneval,51.8,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Gemma 2 27B', rank 63, accessed 2026-04-29). Setting unknown."
gemma-3-1b,humaneval,41.5,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Gemma 3 1B', rank 64, accessed 2026-04-29). Setting unknown."
gemma-2-9b,humaneval,40.2,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Gemma 2 9B', rank 65, accessed 2026-04-29). Setting unknown."
ministral-8b,humaneval,34.8,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Ministral 8B Instruct', rank 66, accessed 2026-04-29). Setting unknown."
minicpm-sala,humaneval,95.1,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'MiniCPM-SALA', rank 1, accessed 2026-04-29). Setting unknown."
qwen2.5-coder-32b-instruct,humaneval,92.7,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Qwen2.5-Coder 32B Instruct', rank 6, accessed 2026-04-29). Setting unknown."
sarvam-30b,humaneval,92.1,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Sarvam-30B', rank 8, accessed 2026-04-29). Setting unknown."
qwen2.5-vl-32b-instruct,humaneval,91.5,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Qwen2.5 VL 32B Instruct', rank 11, accessed 2026-04-29). Setting unknown."
granite-3.3-8b-instruct,humaneval,89.7,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Granite 3.3 8B Instruct', rank 13, accessed 2026-04-29). Setting unknown."
granite-3.3-8b-base,humaneval,89.7,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Granite 3.3 8B Base', rank 13, accessed 2026-04-29). Setting unknown."
gemini-diffusion,humaneval,89.6,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Gemini Diffusion', rank 15, accessed 2026-04-29). Setting unknown."
longcat-flash-chat,humaneval,88.4,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'LongCat-Flash-Chat', rank 19, accessed 2026-04-29). Setting unknown."
qwen2.5-coder-7b-instruct,humaneval,88.4,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Qwen2.5-Coder 7B Instruct', rank 21, accessed 2026-04-29). Setting unknown."
grok-2,humaneval,88.4,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Grok-2', rank 21, accessed 2026-04-29). Setting unknown."
claude-3.5-haiku,humaneval,88.1,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Claude 3.5 Haiku', rank 25, accessed 2026-04-29). Setting unknown."
gpt-4-turbo,humaneval,87.1,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'GPT-4 Turbo', rank 30, accessed 2026-04-29). Setting unknown."
qwen2-72b-instruct,humaneval,86.0,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Qwen2 72B Instruct', rank 32, accessed 2026-04-29). Setting unknown."
grok-2-mini,humaneval,85.7,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Grok-2 mini', rank 33, accessed 2026-04-29). Setting unknown."
nova-lite,humaneval,85.4,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Nova Lite', rank 34, accessed 2026-04-29). Setting unknown."
claude-3-opus,humaneval,84.9,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Claude 3 Opus', rank 36, accessed 2026-04-29). Setting unknown."
mistral-small-3-24b-instruct,humaneval,84.8,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Mistral Small 3 24B Instruct', rank 37, accessed 2026-04-29). Setting unknown."
ibm-granite-4.0-tiny-preview,humaneval,82.4,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'IBM Granite 4.0 Tiny Preview', rank 42, accessed 2026-04-29). Setting unknown."
nova-micro,humaneval,81.1,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Nova Micro', rank 43, accessed 2026-04-29). Setting unknown."
codestral-22b,humaneval,81.1,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Codestral-22B', rank 43, accessed 2026-04-29). Setting unknown."
qwen2-7b-instruct,humaneval,79.9,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Qwen2 7B Instruct', rank 46, accessed 2026-04-29). Setting unknown."
qwen2.5-omni-7b,humaneval,78.7,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Qwen2.5-Omni-7B', rank 47, accessed 2026-04-29). Setting unknown."
claude-3-haiku,humaneval,75.9,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Claude 3 Haiku', rank 48, accessed 2026-04-29). Setting unknown."
gemma-3n-e4b-instructed-litert-preview,humaneval,75.0,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Gemma 3n E4B Instructed LiteRT Preview', rank 49, accessed 2026-04-29). Setting unknown."
gemma-3n-e4b-instructed,humaneval,75.0,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Gemma 3n E4B Instructed', rank 49, accessed 2026-04-29). Setting unknown."
grok-1.5,humaneval,74.1,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Grok-1.5', rank 52, accessed 2026-04-29). Setting unknown."
claude-3-sonnet,humaneval,73.0,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Claude 3 Sonnet', rank 53, accessed 2026-04-29). Setting unknown."
pixtral-12b,humaneval,72.0,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Pixtral-12B', rank 55, accessed 2026-04-29). Setting unknown."
phi-3.5-moe-instruct,humaneval,70.7,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Phi-3.5-MoE-instruct', rank 57, accessed 2026-04-29). Setting unknown."
gemma-3n-e2b-instructed-litert-preview,humaneval,66.5,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Gemma 3n E2B Instructed LiteRT (Preview)', rank 60, accessed 2026-04-29). Setting unknown."
gemma-3n-e2b-instructed,humaneval,66.5,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Gemma 3n E2B Instructed', rank 60, accessed 2026-04-29). Setting unknown."
phi-3.5-mini-instruct,humaneval,62.8,https://llm-stats.com/benchmarks/humaneval,"{""harness"":""unknown (llm-stats aggregator)"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Setting not disclosed by aggregator"",""sampling"":""unknown""}",,third_party_aggregator,verified_third_party,"From llm-stats.com HumanEval leaderboard (display name 'Phi-3.5-mini-instruct', rank 62, accessed 2026-04-29). Setting unknown."
claude-3.5-sonnet,ai2d,94.7,https://llm-stats.com/benchmarks/ai2d,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per DeepSeek V3 paper Table 6."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3.5 Sonnet, slug=claude-3-5-sonnet-20241022, provider=Anthropic"
qwen3.6-plus,ai2d,94.4,https://llm-stats.com/benchmarks/ai2d,{},True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3.6 Plus, slug=qwen3.6-plus, provider=Alibaba Cloud / Qwen Team"
gemma-3-27b,ai2d,84.5,https://llm-stats.com/benchmarks/ai2d,"{""context"":""default (128K)"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemma 3 tech report (arxiv:2503.19786) Table 6 IT column for 27B."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemma 3 27B, slug=gemma-3-27b-it, provider=Google"
gemma-3-12b,ai2d,84.2,https://llm-stats.com/benchmarks/ai2d,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemma 3 tech report Table 6 IT instruction-tuned column."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemma 3 12B, slug=gemma-3-12b-it, provider=Google"
qwen2.5-omni-7b,ai2d,83.2,https://llm-stats.com/benchmarks/ai2d,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5-Omni-7B, slug=qwen2.5-omni-7b, provider=Alibaba Cloud / Qwen Team"
deepseek-v3.2,aider_polyglot_diff,74.5,https://llm-stats.com/benchmarks/aider-polyglot,"{""variant"":""exp""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek-V3.2-Exp, slug=deepseek-v3.2-exp, provider=DeepSeek"
gemini-2.5-flash,aider_polyglot_diff,61.9,https://llm-stats.com/benchmarks/aider-polyglot,"{""context"":""default (1M)"",""effort"":""default"",""harness"":""AI Studio API (gemini-2.5-flash GA)"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Gemini 2.5 Flash Model Card (Dec 2025) GA Thinking column."",""prompt_style"":""default"",""sampling"":""pass@1 (single attempt; multi-trial avg for smaller bench)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 2.5 Flash, slug=gemini-2.5-flash, provider=Google"
qwen3-235b,aider_polyglot_diff,57.3,https://llm-stats.com/benchmarks/aider-polyglot,"{""mode"":""instruct"",""snapshot"":""2507""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3-235B-A22B-Instruct-2507, slug=qwen3-235b-a22b-instruct-2507, provider=Alibaba Cloud / Qwen Team"
gpt-5,aider_polyglot_whole,88.0,https://llm-stats.com/benchmarks/aider-polyglot,"{""context"":""default"",""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per OpenAI GPT-5 dev blog. high reasoning effort."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-5, slug=gpt-5-2025-08-07, provider=OpenAI"
gemini-2.5-pro,aider_polyglot_whole,82.2,https://llm-stats.com/benchmarks/aider-polyglot,"{""context"":""default"",""effort"":""default"",""harness"":""AI Studio API"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Gemini 2.5 Pro Model Card (GA, June 27 2025): pass@1 single attempt, AI Studio API model-id gemini-2.5-pro-preview-06-05 / gemini-2.5-pro GA, default sampling. Release date set to GA date 2025-06-27."",""prompt_style"":""default"",""sampling"":""pass@1 (single attempt)"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 2.5 Pro Preview 06-05, slug=gemini-2.5-pro-preview-06-05, provider=Google"
o3-high,aider_polyglot_whole,81.3,https://llm-stats.com/benchmarks/aider-polyglot,"{""effort"":""default (o3 base, not high)""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=o3, slug=o3-2025-04-16, provider=OpenAI"
deepseek-v3.2,aider_polyglot_whole,74.5,https://llm-stats.com/benchmarks/aider-polyglot,"{""variant"":""exp""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek-V3.2-Exp, slug=deepseek-v3.2-exp, provider=DeepSeek"
deepseek-r1,aider_polyglot_whole,71.6,https://llm-stats.com/benchmarks/aider-polyglot,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per DeepSeek R1-0528 model card."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek-R1-0528, slug=deepseek-r1-0528, provider=DeepSeek"
o4-mini-high,aider_polyglot_whole,68.9,https://llm-stats.com/benchmarks/aider-polyglot,"{""effort"":""default (o4-mini base, not high)""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=o4-mini, slug=o4-mini, provider=OpenAI"
kimi-k2,aider_polyglot_whole,60.0,https://llm-stats.com/benchmarks/aider-polyglot,"{""snapshot"":""0905"",""variant"":""instruct""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Kimi K2-Instruct-0905, slug=kimi-k2-instruct-0905, provider=Moonshot AI"
qwen3-235b,aider_polyglot_whole,57.3,https://llm-stats.com/benchmarks/aider-polyglot,"{""mode"":""instruct"",""snapshot"":""2507""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3-235B-A22B-Instruct-2507, slug=qwen3-235b-a22b-instruct-2507, provider=Alibaba Cloud / Qwen Team"
deepseek-v3,aider_polyglot_whole,49.6,https://llm-stats.com/benchmarks/aider-polyglot,{},True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek-V3, slug=deepseek-v3, provider=DeepSeek"
ministral-8b,aime_2024,86.0,https://llm-stats.com/benchmarks/aime-2024,"{""context"":""default"",""effort"":""n/a"",""harness"":""OLMES (Allen AI)"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OLMo 2 paper Table 7 Instruct (instruction-tuned variant)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Ministral 3 (8B Reasoning 2512), slug=ministral-8b-latest, provider=Mistral AI"
minicpm-sala,aime_2024,83.8,https://llm-stats.com/benchmarks/aime-2024,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=MiniCPM-SALA, slug=minicpm-sala, provider=OpenBMB"
granite-3.3-8b-base,aime_2024,81.2,https://llm-stats.com/benchmarks/aime-2024,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Granite 3.3 8B Base, slug=granite-3.3-8b-base, provider=IBM"
granite-3.3-8b-instruct,aime_2024,81.2,https://llm-stats.com/benchmarks/aime-2024,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Granite 3.3 8B Instruct, slug=granite-3.3-8b-instruct, provider=IBM"
o4-mini-high,mt_aime_2024,93.4,https://llm-stats.com/benchmarks/aime-2024,"{""effort"":""default (o4-mini base, not high)""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=o4-mini, slug=o4-mini, provider=OpenAI"
grok-3-beta,mt_aime_2024,93.3,https://llm-stats.com/benchmarks/aime-2024,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per xAI Grok 3 blog: canonical = non-reasoning mode (the cross-model table without 'Think' label). Reasoning ON values not used."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Grok-3, slug=grok-3, provider=xAI"
gemini-2.5-pro,mt_aime_2024,92.0,https://llm-stats.com/benchmarks/aime-2024,"{""context"":""default"",""effort"":""default"",""harness"":""AI Studio API"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Gemini 2.5 Pro Model Card (GA, June 27 2025): pass@1 single attempt, AI Studio API model-id gemini-2.5-pro-preview-06-05 / gemini-2.5-pro GA, default sampling. Release date set to GA date 2025-06-27."",""prompt_style"":""default"",""sampling"":""pass@1 (single attempt)"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 2.5 Pro, slug=gemini-2.5-pro, provider=Google"
o3-high,mt_aime_2024,91.6,https://llm-stats.com/benchmarks/aime-2024,"{""effort"":""default (o3 base, not high)""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=o3, slug=o3-2025-04-16, provider=OpenAI"
deepseek-r1,mt_aime_2024,91.4,https://llm-stats.com/benchmarks/aime-2024,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per DeepSeek R1-0528 model card."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek-R1-0528, slug=deepseek-r1-0528, provider=DeepSeek"
gemini-2.5-flash,mt_aime_2024,88.0,https://llm-stats.com/benchmarks/aime-2024,"{""context"":""default (1M)"",""effort"":""default"",""harness"":""AI Studio API (gemini-2.5-flash GA)"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Gemini 2.5 Flash Model Card (Dec 2025) GA Thinking column."",""prompt_style"":""default"",""sampling"":""pass@1 (single attempt; multi-trial avg for smaller bench)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 2.5 Flash, slug=gemini-2.5-flash, provider=Google"
ministral-8b,mt_aime_2024,86.0,https://llm-stats.com/benchmarks/aime-2024,"{""context"":""default"",""effort"":""n/a"",""harness"":""OLMES (Allen AI)"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OLMo 2 paper Table 7 Instruct (instruction-tuned variant)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Ministral 3 (8B Reasoning 2512), slug=ministral-8b-latest, provider=Mistral AI"
minicpm-sala,mt_aime_2024,83.8,https://llm-stats.com/benchmarks/aime-2024,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=MiniCPM-SALA, slug=minicpm-sala, provider=OpenBMB"
deepseek-r1-distill-qwen-32b,mt_aime_2024,83.3,https://llm-stats.com/benchmarks/aime-2024,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per DeepSeek R1 paper Table 15 (distilled model)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek R1 Distill Qwen 32B, slug=deepseek-r1-distill-qwen-32b, provider=DeepSeek"
deepseek-r1-distill-qwen-7b,mt_aime_2024,83.3,https://llm-stats.com/benchmarks/aime-2024,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per DeepSeek R1 paper Table 15 (distilled model)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek R1 Distill Qwen 7B, slug=deepseek-r1-distill-qwen-7b, provider=DeepSeek"
phi-4-reasoning-plus,mt_aime_2024,81.3,https://llm-stats.com/benchmarks/aime-2024,"{""context"":""default (32K)"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Phi-4-reasoning paper. Same as phi-4-reasoning + GRPO RL on top."",""prompt_style"":""default"",""sampling"":""pass@1 (avg of 5)"",""temperature"":""0.8"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Phi 4 Reasoning Plus, slug=phi-4-reasoning-plus, provider=Microsoft"
granite-3.3-8b-base,mt_aime_2024,81.2,https://llm-stats.com/benchmarks/aime-2024,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Granite 3.3 8B Base, slug=granite-3.3-8b-base, provider=IBM"
granite-3.3-8b-instruct,mt_aime_2024,81.2,https://llm-stats.com/benchmarks/aime-2024,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Granite 3.3 8B Instruct, slug=granite-3.3-8b-instruct, provider=IBM"
deepseek-r1-distill-llama-8b,mt_aime_2024,80.0,https://llm-stats.com/benchmarks/aime-2024,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per DeepSeek R1 paper Table 15 (distilled model)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek R1 Distill Llama 8B, slug=deepseek-r1-distill-llama-8b, provider=DeepSeek"
deepseek-r1-distill-qwen-14b,mt_aime_2024,80.0,https://llm-stats.com/benchmarks/aime-2024,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per DeepSeek R1 paper Table 15 (distilled model)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek R1 Distill Qwen 14B, slug=deepseek-r1-distill-qwen-14b, provider=DeepSeek"
claude-3.7-sonnet,mt_aime_2024,80.0,https://llm-stats.com/benchmarks/aime-2024,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Claude 4 blog cross-model table: with extended thinking for GPQA/MMMU/AIME/TAU; no extended thinking for SWE/Terminal/MMMLU."",""prompt_style"":""zeroshot-cot"",""sampling"":""pass@1"",""temperature"":0.0,""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3.7 Sonnet, slug=claude-3-7-sonnet-20250219, provider=Anthropic"
phi-4-reasoning,mt_aime_2024,75.3,https://llm-stats.com/benchmarks/aime-2024,"{""context"":""default (32K)"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Phi-4-reasoning paper (arxiv:2504.21318) Tables 1-2. Thinking model, temp 0.8, max_seq_len 31k."",""prompt_style"":""default"",""sampling"":""pass@1 (avg of 5)"",""temperature"":""0.8"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Phi 4 Reasoning, slug=phi-4-reasoning, provider=Microsoft"
o1-high,mt_aime_2024,74.3,https://llm-stats.com/benchmarks/aime-2024,"{""effort"":""default (o1 base, not high)""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=o1, slug=o1-2024-12-17, provider=OpenAI"
gemini-2.0-flash,mt_aime_2024,73.3,https://llm-stats.com/benchmarks/aime-2024,"{""effort"":""thinking""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 2.0 Flash Thinking, slug=gemini-2.0-flash-thinking, provider=Google"
kimi-k2,mt_aime_2024,72.0,https://llm-stats.com/benchmarks/aime-2024,"{""context"":""default (256K)"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Kimi K2 tech report (arxiv:2507.20534) Table 3 Kimi-K2-Instruct column. Non-thinking model."",""prompt_style"":""default"",""sampling"":""pass@1 (avg of 4-64 trials per benchmark)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Kimi K2 0905, slug=kimi-k2-0905, provider=Moonshot AI"
deepseek-v3,mt_aime_2024,59.4,https://llm-stats.com/benchmarks/aime-2024,{},True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek-V3 0324, slug=deepseek-v3-0324, provider=DeepSeek"
deepseek-r1-distill-qwen-1.5b,mt_aime_2024,52.7,https://llm-stats.com/benchmarks/aime-2024,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per DeepSeek R1 paper Table 15 (distilled model)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek R1 Distill Qwen 1.5B, slug=deepseek-r1-distill-qwen-1.5b, provider=DeepSeek"
gpt-4.1-mini,mt_aime_2024,49.6,https://llm-stats.com/benchmarks/aime-2024,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OpenAI GPT-5 dev blog. non-reasoning model."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-4.1 mini, slug=gpt-4.1-mini-2025-04-14, provider=OpenAI"
gpt-4.1,mt_aime_2024,48.1,https://llm-stats.com/benchmarks/aime-2024,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OpenAI GPT-5 dev blog. non-reasoning model."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-4.1, slug=gpt-4.1-2025-04-14, provider=OpenAI"
gpt-4.5,mt_aime_2024,36.7,https://llm-stats.com/benchmarks/aime-2024,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OpenAI GPT-4.1 blog table."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-4.5, slug=gpt-4.5, provider=OpenAI"
gpt-4.1-nano,mt_aime_2024,29.4,https://llm-stats.com/benchmarks/aime-2024,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OpenAI GPT-5 dev blog. non-reasoning model."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-4.1 nano, slug=gpt-4.1-nano-2025-04-14, provider=OpenAI"
gpt-4o,mt_aime_2024,13.1,https://llm-stats.com/benchmarks/aime-2024,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OpenAI GPT-4.1 blog table."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-4o, slug=gpt-4o-2024-08-06, provider=OpenAI"
sarvam-30b,aime_2025,96.7,https://llm-stats.com/benchmarks/aime-2025,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Sarvam-30B, slug=sarvam-30b, provider=Sarvam AI"
ministral-8b,aime_2025,78.7,https://llm-stats.com/benchmarks/aime-2025,"{""context"":""default"",""effort"":""n/a"",""harness"":""OLMES (Allen AI)"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OLMo 2 paper Table 7 Instruct (instruction-tuned variant)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Ministral 3 (8B Reasoning 2512), slug=ministral-8b-latest, provider=Mistral AI"
minicpm-sala,aime_2025,78.3,https://llm-stats.com/benchmarks/aime-2025,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=MiniCPM-SALA, slug=minicpm-sala, provider=OpenBMB"
longcat-flash-chat,aime_2025,61.3,https://llm-stats.com/benchmarks/aime-2025,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=LongCat-Flash-Chat, slug=longcat-flash-chat, provider=Meituan"
gemini-diffusion,aime_2025,23.3,https://llm-stats.com/benchmarks/aime-2025,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini Diffusion, slug=gemini-diffusion, provider=Google"
mistral-small-24b,arena_hard,87.6,https://llm-stats.com/benchmarks/arena-hard,"{""context"":""default"",""effort"":""n/a"",""harness"":""OLMES (Allen AI)"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OLMo 2 paper Table 7 Instruct (instruction-tuned variant)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Mistral Small 3 24B Instruct, slug=mistral-small-24b-instruct-2501, provider=Mistral AI"
deepseek-v2.5-0905,arena_hard,76.2,https://llm-stats.com/benchmarks/arena-hard,"{""snapshot"":""0905""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek-V2.5, slug=deepseek-v2.5, provider=DeepSeek"
granite-3.3-8b-base,arena_hard,57.6,https://llm-stats.com/benchmarks/arena-hard,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Granite 3.3 8B Base, slug=granite-3.3-8b-base, provider=IBM"
granite-3.3-8b-instruct,arena_hard,57.6,https://llm-stats.com/benchmarks/arena-hard,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Granite 3.3 8B Instruct, slug=granite-3.3-8b-instruct, provider=IBM"
mistral-large-3,arena_hard,66.7,https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/resolve/a11f36bebf709121056b1dbcc943d1c6afbe494d/images/image3.png?download=true,"{""context"":""source does not state"",""effort"":""none"",""harness"":""Mistral official model-card evaluation"",""judge"":""Arena-Hard automatic judge"",""mode"":""non-thinking"",""prompt_style"":""official/default or source does not state"",""sampling"":""pass@1; repeat count undisclosed"",""temperature"":""source does not state"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly: '66.7'. Research observation: small-image3.png:4:4:score:reported.
qwen-2.5-7b,arena_hard,52.0,https://llm-stats.com/benchmarks/arena-hard,"{""context"":""default"",""effort"":""n/a"",""harness"":""OLMES (Allen AI)"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OLMo 2 paper Table 7 Instruct (instruction-tuned variant)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5 7B Instruct, slug=qwen-2.5-7b-instruct, provider=Alibaba Cloud / Qwen Team"
phi-3.5-moe-instruct,arena_hard,37.9,https://llm-stats.com/benchmarks/arena-hard,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Phi-3.5-MoE-instruct, slug=phi-3.5-moe-instruct, provider=Microsoft"
phi-3.5-mini-instruct,arena_hard,37.0,https://llm-stats.com/benchmarks/arena-hard,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Phi-3.5-mini-instruct, slug=phi-3.5-mini-instruct, provider=Microsoft"
phi-4-mini,arena_hard,32.8,https://llm-stats.com/benchmarks/arena-hard,"{""context"":""default (128K)"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Phi-4-mini-instruct HF model card."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Phi 4 Mini, slug=phi-4-mini, provider=Microsoft"
sarvam-30b,beyond_aime,58.3,https://llm-stats.com/benchmarks/beyond-aime,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Sarvam-30B, slug=sarvam-30b, provider=Sarvam AI"
minimax-m2.5,bfcl_v3_multiturn,76.8,https://llm-stats.com/benchmarks/bfcl-v3-multiturn,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per MiniMax M2.5 model card."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=MiniMax M2.5, slug=minimax-m2.5, provider=MiniMax"
llama-3.1-405b-instruct,bfcl,88.5,https://llm-stats.com/benchmarks/bfcl,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per DeepSeek V3 paper Table 6."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Llama 3.1 405B Instruct, slug=llama-3.1-405b-instruct, provider=Meta"
llama-3.1-70b,bfcl,84.8,https://llm-stats.com/benchmarks/bfcl,"{""context"":""default"",""effort"":""n/a"",""harness"":""OLMES (Allen AI)"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OLMo 2 paper Table 7 Instruct (instruction-tuned variant)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Llama 3.1 70B Instruct, slug=llama-3.1-70b-instruct, provider=Meta"
qwen3-235b,bfcl,70.8,https://llm-stats.com/benchmarks/bfcl,"{""context"":""default"",""effort"":""default"",""harness"":""official (Qwen-Agent / sglang)"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Qwen3 tech report (arxiv:2505.09388) Table 11 Thinking column. Output length 32768 (38912 for AIME)."",""prompt_style"":""default (enable_thinking=True)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3 235B A22B, slug=qwen3-235b-a22b, provider=Alibaba Cloud / Qwen Team"
qwen3-32b,bfcl,70.3,https://llm-stats.com/benchmarks/bfcl,"{""context"":""default"",""effort"":""default"",""harness"":""official (Qwen-Agent / sglang)"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Qwen3 tech report (arxiv:2505.09388) Table 13/15/17/19 Thinking column."",""prompt_style"":""default (enable_thinking=True)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3 32B, slug=qwen3-32b, provider=Alibaba Cloud / Qwen Team"
qwen3-30b-a3b,bfcl,69.1,https://llm-stats.com/benchmarks/bfcl,"{""context"":""default"",""effort"":""default"",""harness"":""official (Qwen-Agent / sglang)"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Qwen3 tech report (arxiv:2505.09388) Table 13/15/17/19 Thinking column."",""prompt_style"":""default (enable_thinking=True)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3 30B A3B, slug=qwen3-30b-a3b, provider=Alibaba Cloud / Qwen Team"
nova-lite,bfcl,66.6,https://llm-stats.com/benchmarks/bfcl,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Nova Lite, slug=nova-lite, provider=Amazon"
qwq-32b,bfcl,66.4,https://llm-stats.com/benchmarks/bfcl,"{""context"":""default (32K)"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per QwQ-32B release blog (qwenlm.github.io/blog/qwq-32b/) + Qwen3 tech report Table 13 baseline column."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=QwQ-32B, slug=qwq-32b, provider=Alibaba Cloud / Qwen Team"
nova-micro,bfcl,56.2,https://llm-stats.com/benchmarks/bfcl,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Nova Micro, slug=nova-micro, provider=Amazon"
gemma-3-27b,bigbench_extra_hard,19.3,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-3-12b,bigbench_extra_hard,16.3,https://llm-stats.com/benchmarks/big-bench-extra-hard,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemma 3 tech report Table 6 IT instruction-tuned column."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemma 3 12B, slug=gemma-3-12b-it, provider=Google"
gemini-diffusion,bigbench_extra_hard,15.0,https://llm-stats.com/benchmarks/big-bench-extra-hard,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini Diffusion, slug=gemini-diffusion, provider=Google"
gemma-3-1b-it,bigbench_extra_hard,7.2,https://llm-stats.com/benchmarks/big-bench-extra-hard,"{""context"":""32k"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Google Gemma 3 1B instruction-tuned."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemma 3 1B, slug=gemma-3-1b-it, provider=Google"
claude-3.5-sonnet,bigbench_hard,93.1,https://llm-stats.com/benchmarks/big-bench-hard,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per DeepSeek V3 paper Table 6."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3.5 Sonnet, slug=claude-3-5-sonnet-20241022, provider=Anthropic"
gemini-1.5-pro,bigbench_hard,89.2,https://llm-stats.com/benchmarks/big-bench-hard,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemini 2.0 Flash Model Card (April 15 2025): non-thinking generation, default sampling. Same canonical setting applies to 1.5 / 2.0 family."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 1.5 Pro, slug=gemini-1.5-pro, provider=Google"
claude-3-opus,bigbench_hard,86.8,https://llm-stats.com/benchmarks/big-bench-hard,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3 Opus, slug=claude-3-opus-20240229, provider=Anthropic"
gemma-3-12b,bigbench_hard,85.7,https://llm-stats.com/benchmarks/big-bench-hard,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemma 3 tech report Table 6 IT instruction-tuned column."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemma 3 12B, slug=gemma-3-12b-it, provider=Google"
gemini-1.5-flash,bigbench_hard,85.5,https://llm-stats.com/benchmarks/big-bench-hard,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemini 2.0 Flash Model Card (April 15 2025): non-thinking generation, default sampling. Same canonical setting applies to 1.5 / 2.0 family."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 1.5 Flash, slug=gemini-1.5-flash, provider=Google"
claude-3-sonnet,bigbench_hard,82.9,https://llm-stats.com/benchmarks/big-bench-hard,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3 Sonnet, slug=claude-3-sonnet-20240229, provider=Anthropic"
phi-3.5-moe-instruct,bigbench_hard,79.1,https://llm-stats.com/benchmarks/big-bench-hard,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Phi-3.5-MoE-instruct, slug=phi-3.5-moe-instruct, provider=Microsoft"
claude-3-haiku,bigbench_hard,73.7,https://llm-stats.com/benchmarks/big-bench-hard,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3 Haiku, slug=claude-3-haiku-20240307, provider=Anthropic"
granite-3.3-8b-instruct,bigbench_hard,69.1,https://llm-stats.com/benchmarks/big-bench-hard,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Granite 3.3 8B Instruct, slug=granite-3.3-8b-instruct, provider=IBM"
granite-3.3-8b-base,bigbench_hard,69.1,https://llm-stats.com/benchmarks/big-bench-hard,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Granite 3.3 8B Base, slug=granite-3.3-8b-base, provider=IBM"
phi-3.5-mini-instruct,bigbench_hard,69.0,https://llm-stats.com/benchmarks/big-bench-hard,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Phi-3.5-mini-instruct, slug=phi-3.5-mini-instruct, provider=Microsoft"
gemma-3-1b-it,bigbench_hard,39.1,https://llm-stats.com/benchmarks/big-bench-hard,"{""context"":""32k"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Google Gemma 3 1B instruction-tuned."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemma 3 1B, slug=gemma-3-1b-it, provider=Google"
gemini-diffusion,bigcodebench,45.4,https://llm-stats.com/benchmarks/bigcodebench,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini Diffusion, slug=gemini-diffusion, provider=Google"
gemma-3-1b-it,bird_sql,6.4,https://llm-stats.com/benchmarks/bird-sql-(dev),"{""context"":""32k"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Google Gemma 3 1B instruction-tuned."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemma 3 1B, slug=gemma-3-1b-it, provider=Google"
gpt-5.2,browsecomp_long_context_128k,92.0,https://llm-stats.com/benchmarks/browsecomp-long-128k,"{""context"":""default"",""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per OpenAI gpt-5.2 blog: reasoning effort=xhigh, research environment."",""prompt_style"":""default (with prompt adjustment for \u03c4\u00b2-bench)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-5.2, slug=gpt-5.2-2025-12-11, provider=OpenAI"
gpt-5.2,browsecomp_long_context_256k,89.8,https://llm-stats.com/benchmarks/browsecomp-long-256k,"{""context"":""default"",""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per OpenAI gpt-5.2 blog: reasoning effort=xhigh, research environment."",""prompt_style"":""default (with prompt adjustment for \u03c4\u00b2-bench)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-5.2, slug=gpt-5.2-2025-12-11, provider=OpenAI"
gpt-5,browsecomp_long_context_256k,88.8,https://llm-stats.com/benchmarks/browsecomp-long-256k,"{""context"":""default"",""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per OpenAI GPT-5 dev blog. high reasoning effort."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-5, slug=gpt-5-2025-08-07, provider=OpenAI"
deepseek-r1,browsecomp_zh,35.7,https://llm-stats.com/benchmarks/browsecomp-zh,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per DeepSeek R1-0528 model card."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek-R1-0528, slug=deepseek-r1-0528, provider=DeepSeek"
minimax-m2.5,browsecomp,76.3,https://llm-stats.com/benchmarks/browsecomp,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per MiniMax M2.5 model card."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=MiniMax M2.5, slug=minimax-m2.5, provider=MiniMax"
minimax-m2.1,browsecomp,62.0,https://llm-stats.com/benchmarks/browsecomp,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per MiniMax M2.5 model card (referenced as predecessor)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=MiniMax M2.1, slug=minimax-m2.1, provider=MiniMax"
o4-mini-high,browsecomp,51.5,https://llm-stats.com/benchmarks/browsecomp,"{""effort"":""default (o4-mini base, not high)""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=o4-mini, slug=o4-mini, provider=OpenAI"
sarvam-30b,browsecomp,35.5,https://llm-stats.com/benchmarks/browsecomp,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Sarvam-30B, slug=sarvam-30b, provider=Sarvam AI"
deepseek-r1,browsecomp,8.9,https://llm-stats.com/benchmarks/browsecomp,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per DeepSeek R1-0528 model card."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek-R1-0528, slug=deepseek-r1-0528, provider=DeepSeek"
qwen3.6-plus,c_eval,93.3,https://llm-stats.com/benchmarks/c-eval,{},True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3.6 Plus, slug=qwen3.6-plus, provider=Alibaba Cloud / Qwen Team"
qwen2-72b-instruct,c_eval,83.8,https://llm-stats.com/benchmarks/c-eval,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2 72B Instruct, slug=qwen2-72b-instruct, provider=Alibaba Cloud / Qwen Team"
qwen2-7b-instruct,c_eval,77.2,https://llm-stats.com/benchmarks/c-eval,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2 7B Instruct, slug=qwen2-7b-instruct, provider=Alibaba Cloud / Qwen Team"
claude-3.5-sonnet,chartqa,90.8,https://llm-stats.com/benchmarks/chartqa,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per DeepSeek V3 paper Table 6."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3.5 Sonnet, slug=claude-3-5-sonnet-20241022, provider=Anthropic"
llama-4-scout,chartqa,88.8,https://llm-stats.com/benchmarks/chartqa,"{""context"":""default"",""effort"":""n/a"",""harness"":""official-meta"",""judge"":""rule-based"",""mode"":""non-thinking"",""multimodal_input"":true,""prompt_style"":""zeroshot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Llama 4 Scout, slug=llama-4-scout, provider=Meta"
nova-lite,chartqa,86.8,https://llm-stats.com/benchmarks/chartqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Nova Lite, slug=nova-lite, provider=Amazon"
qwen2.5-omni-7b,chartqa,85.3,https://llm-stats.com/benchmarks/chartqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5-Omni-7B, slug=qwen2.5-omni-7b, provider=Alibaba Cloud / Qwen Team"
pixtral-12b,chartqa,81.8,https://llm-stats.com/benchmarks/chartqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Pixtral-12B, slug=pixtral-12b-2409, provider=Mistral AI"
gemma-3-27b,chartqa,78.0,https://llm-stats.com/benchmarks/chartqa,"{""context"":""default (128K)"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemma 3 tech report (arxiv:2503.19786) Table 6 IT column for 27B."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemma 3 27B, slug=gemma-3-27b-it, provider=Google"
gemma-3-12b,chartqa,75.7,https://llm-stats.com/benchmarks/chartqa,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemma 3 tech report Table 6 IT instruction-tuned column."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemma 3 12B, slug=gemma-3-12b-it, provider=Google"
qwen2-72b-instruct,cmmlu,90.1,https://llm-stats.com/benchmarks/cmmlu,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2 72B Instruct, slug=qwen2-72b-instruct, provider=Alibaba Cloud / Qwen Team"
longcat-flash-chat,cmmlu,84.3,https://llm-stats.com/benchmarks/cmmlu,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=LongCat-Flash-Chat, slug=longcat-flash-chat, provider=Meituan"
minicpm-sala,cmmlu,81.6,https://llm-stats.com/benchmarks/cmmlu,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=MiniCPM-SALA, slug=minicpm-sala, provider=OpenBMB"
deepseek-v4-pro,codeforces_avg8,100.0,https://llm-stats.com/benchmarks/codeforces,"{""effort"":""max""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek-V4-Pro-Max, slug=deepseek-v4-pro-max, provider=DeepSeek"
deepseek-v4-flash,codeforces_avg8,100.0,https://llm-stats.com/benchmarks/codeforces,"{""effort"":""max""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek-V4-Flash-Max, slug=deepseek-v4-flash-max, provider=DeepSeek"
deepseek-v3.2-speciale,codeforces_avg8,90.0,https://llm-stats.com/benchmarks/codeforces,"{""context"":""128K"",""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""DS V3.2-Speciale: high-compute reasoning variant of V3.2. Per V3.2 tech report Table 3."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek-V3.2-Speciale, slug=deepseek-v3.2-speciale, provider=DeepSeek"
gpt-oss-120b,codeforces_avg8,82.1,https://llm-stats.com/benchmarks/codeforces,"{""context"":""default (128K)"",""effort"":""high"",""harness"":""official (harmony chat format)"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per OpenAI gpt-oss model card (arxiv:2508.10925) Table 3, high reasoning effort. Default with-tools for agentic; AIME/GPQA also has no-tools variant in same table."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""with tools (per-bench override; harmony chat format default)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT OSS 120B, slug=gpt-oss-120b, provider=OpenAI"
deepseek-v3.2,codeforces_avg8,79.5,https://llm-stats.com/benchmarks/codeforces,"{""mode"":""thinking (DeepSeek-Reasoner endpoint)""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek-V3.2 (Thinking), slug=deepseek-reasoner, provider=DeepSeek"
gpt-oss-20b,codeforces_avg8,74.3,https://llm-stats.com/benchmarks/codeforces,"{""context"":""default (128K)"",""effort"":""high"",""harness"":""official (harmony chat format)"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per OpenAI gpt-oss model card (arxiv:2508.10925) Table 3, high reasoning effort. Default with-tools for agentic; AIME/GPQA also has no-tools variant in same table."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""with tools (per-bench override; harmony chat format default)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT OSS 20B, slug=gpt-oss-20b, provider=OpenAI"
qwen3-32b,codeforces_avg8,65.9,https://llm-stats.com/benchmarks/codeforces,"{""context"":""default"",""effort"":""default"",""harness"":""official (Qwen-Agent / sglang)"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Qwen3 tech report (arxiv:2505.09388) Table 13/15/17/19 Thinking column."",""prompt_style"":""default (enable_thinking=True)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3 32B, slug=qwen3-32b, provider=Alibaba Cloud / Qwen Team"
deepseek-v4-pro,codeforces_pass8,100.0,https://llm-stats.com/benchmarks/codeforces,"{""effort"":""max""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek-V4-Pro-Max, slug=deepseek-v4-pro-max, provider=DeepSeek"
deepseek-v4-flash,codeforces_pass8,100.0,https://llm-stats.com/benchmarks/codeforces,"{""effort"":""max""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek-V4-Flash-Max, slug=deepseek-v4-flash-max, provider=DeepSeek"
deepseek-v3.2-speciale,codeforces_pass8,90.0,https://llm-stats.com/benchmarks/codeforces,"{""context"":""128K"",""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""DS V3.2-Speciale: high-compute reasoning variant of V3.2. Per V3.2 tech report Table 3."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek-V3.2-Speciale, slug=deepseek-v3.2-speciale, provider=DeepSeek"
gpt-oss-120b,codeforces_pass8,82.1,https://llm-stats.com/benchmarks/codeforces,"{""context"":""default (128K)"",""effort"":""high"",""harness"":""official (harmony chat format)"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per OpenAI gpt-oss model card (arxiv:2508.10925) Table 3, high reasoning effort. Default with-tools for agentic; AIME/GPQA also has no-tools variant in same table."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""with tools (per-bench override; harmony chat format default)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT OSS 120B, slug=gpt-oss-120b, provider=OpenAI"
deepseek-v3.2,codeforces_pass8,79.5,https://llm-stats.com/benchmarks/codeforces,"{""mode"":""thinking (DeepSeek-Reasoner endpoint)""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek-V3.2 (Thinking), slug=deepseek-reasoner, provider=DeepSeek"
gpt-oss-20b,codeforces_pass8,74.3,https://llm-stats.com/benchmarks/codeforces,"{""context"":""default (128K)"",""effort"":""high"",""harness"":""official (harmony chat format)"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per OpenAI gpt-oss model card (arxiv:2508.10925) Table 3, high reasoning effort. Default with-tools for agentic; AIME/GPQA also has no-tools variant in same table."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""with tools (per-bench override; harmony chat format default)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT OSS 20B, slug=gpt-oss-20b, provider=OpenAI"
qwen3-32b,codeforces_pass8,65.9,https://llm-stats.com/benchmarks/codeforces,"{""context"":""default"",""effort"":""default"",""harness"":""official (Qwen-Agent / sglang)"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Qwen3 tech report (arxiv:2505.09388) Table 13/15/17/19 Thinking column."",""prompt_style"":""default (enable_thinking=True)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3 32B, slug=qwen3-32b, provider=Alibaba Cloud / Qwen Team"
qwen3.6-plus,countbench,97.6,https://llm-stats.com/benchmarks/countbench,{},True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3.6 Plus, slug=qwen3.6-plus, provider=Alibaba Cloud / Qwen Team"
qwen3-235b,creative_writing_v3,87.5,https://llm-stats.com/benchmarks/creative-writing-v3,"{""mode"":""instruct"",""snapshot"":""2507""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3-235B-A22B-Instruct-2507, slug=qwen3-235b-a22b-instruct-2507, provider=Alibaba Cloud / Qwen Team"
claude-mythos,cybench,100.0,https://llm-stats.com/benchmarks/cybench,"{""context"":""up to 1M"",""effort"":""adaptive thinking at max effort"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Mythos System Card Table 6.3.A: standard config is adaptive thinking max effort, default sampling (temperature, top_p), avg 5 trials, context up to 1M."",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude Mythos Preview, slug=claude-mythos-preview, provider=Anthropic"
grok-4.1,cybench,39.0,https://llm-stats.com/benchmarks/cybench,"{""effort"":""thinking""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Grok-4.1 Thinking, slug=grok-4.1-thinking-2025-11-17, provider=xAI"
claude-3.5-sonnet,docvqa,95.2,https://llm-stats.com/benchmarks/docvqa,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per DeepSeek V3 paper Table 6."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3.5 Sonnet, slug=claude-3-5-sonnet-20241022, provider=Anthropic"
qwen2.5-omni-7b,docvqa,95.2,https://llm-stats.com/benchmarks/docvqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5-Omni-7B, slug=qwen2.5-omni-7b, provider=Alibaba Cloud / Qwen Team"
llama-4-scout,docvqa,94.4,https://llm-stats.com/benchmarks/docvqa,"{""context"":""default"",""effort"":""n/a"",""harness"":""official-meta"",""judge"":""rule-based"",""mode"":""non-thinking"",""multimodal_input"":true,""prompt_style"":""zeroshot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Llama 4 Scout, slug=llama-4-scout, provider=Meta"
grok-2,docvqa,93.6,https://llm-stats.com/benchmarks/docvqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Grok-2, slug=grok-2, provider=xAI"
grok-2-mini,docvqa,93.2,https://llm-stats.com/benchmarks/docvqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Grok-2 mini, slug=grok-2-mini, provider=xAI"
nova-lite,docvqa,92.4,https://llm-stats.com/benchmarks/docvqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Nova Lite, slug=nova-lite, provider=Amazon"
pixtral-12b,docvqa,90.7,https://llm-stats.com/benchmarks/docvqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Pixtral-12B, slug=pixtral-12b-2409, provider=Mistral AI"
gemma-3-12b,docvqa,87.1,https://llm-stats.com/benchmarks/docvqa,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemma 3 tech report Table 6 IT instruction-tuned column."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemma 3 12B, slug=gemma-3-12b-it, provider=Google"
gemma-3-27b,docvqa,86.6,https://llm-stats.com/benchmarks/docvqa,"{""context"":""default (128K)"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemma 3 tech report (arxiv:2503.19786) Table 6 IT column for 27B."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemma 3 27B, slug=gemma-3-27b-it, provider=Google"
grok-1.5,docvqa,85.6,https://llm-stats.com/benchmarks/docvqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Grok-1.5, slug=grok-1.5, provider=xAI"
gpt-4-turbo,drop,86.0,https://llm-stats.com/benchmarks/drop,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-4 Turbo, slug=gpt-4-turbo-2024-04-09, provider=OpenAI"
gpt-4o,drop,83.4,https://llm-stats.com/benchmarks/drop,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OpenAI GPT-4.1 blog table."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-4o, slug=gpt-4o-2024-05-13, provider=OpenAI"
claude-3.5-haiku,drop,83.1,https://llm-stats.com/benchmarks/drop,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3.5 Haiku, slug=claude-3-5-haiku-20241022, provider=Anthropic"
claude-3-opus,drop,83.1,https://llm-stats.com/benchmarks/drop,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3 Opus, slug=claude-3-opus-20240229, provider=Anthropic"
nova-lite,drop,80.2,https://llm-stats.com/benchmarks/drop,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Nova Lite, slug=nova-lite, provider=Amazon"
nova-micro,drop,79.3,https://llm-stats.com/benchmarks/drop,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Nova Micro, slug=nova-micro, provider=Amazon"
longcat-flash-chat,drop,79.1,https://llm-stats.com/benchmarks/drop,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=LongCat-Flash-Chat, slug=longcat-flash-chat, provider=Meituan"
claude-3-sonnet,drop,78.9,https://llm-stats.com/benchmarks/drop,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3 Sonnet, slug=claude-3-sonnet-20240229, provider=Anthropic"
claude-3-haiku,drop,78.4,https://llm-stats.com/benchmarks/drop,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3 Haiku, slug=claude-3-haiku-20240307, provider=Anthropic"
gemini-1.5-pro,drop,74.9,https://llm-stats.com/benchmarks/drop,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemini 2.0 Flash Model Card (April 15 2025): non-thinking generation, default sampling. Same canonical setting applies to 1.5 / 2.0 family."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 1.5 Pro, slug=gemini-1.5-pro, provider=Google"
granite-3.3-8b-instruct,drop,59.4,https://llm-stats.com/benchmarks/drop,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Granite 3.3 8B Instruct, slug=granite-3.3-8b-instruct, provider=IBM"
granite-3.3-8b-base,drop,36.1,https://llm-stats.com/benchmarks/drop,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Granite 3.3 8B Base, slug=granite-3.3-8b-base, provider=IBM"
qwen3.6-plus,dynamath,88.0,https://llm-stats.com/benchmarks/dynamath,{},True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3.6 Plus, slug=qwen3.6-plus, provider=Alibaba Cloud / Qwen Team"
gpt-4o,egoschema,72.2,https://llm-stats.com/benchmarks/egoschema,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OpenAI GPT-4.1 blog table."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-4o, slug=gpt-4o-2024-08-06, provider=OpenAI"
nova-lite,egoschema,71.4,https://llm-stats.com/benchmarks/egoschema,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Nova Lite, slug=nova-lite, provider=Amazon"
qwen2.5-omni-7b,egoschema,68.6,https://llm-stats.com/benchmarks/egoschema,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5-Omni-7B, slug=qwen2.5-omni-7b, provider=Alibaba Cloud / Qwen Team"
qwen3.6-plus,erqa,65.7,https://llm-stats.com/benchmarks/erqa,{},True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3.6 Plus, slug=qwen3.6-plus, provider=Alibaba Cloud / Qwen Team"
gpt-4o,erqa,35.2,https://llm-stats.com/benchmarks/erqa,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OpenAI GPT-4.1 blog table."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-4o, slug=gpt-4o-2024-08-06, provider=OpenAI"
gemini-3-pro,facts_grounding,70.5,https://llm-stats.com/benchmarks/facts-grounding,"{""context"":""default"",""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Google DeepMind /models/gemini/pro/ table: Gemini 3 Pro Thinking (High), no tools by default."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 3 Pro, slug=gemini-3-pro-preview, provider=Google"
gemini-3-flash,facts_grounding,61.9,https://llm-stats.com/benchmarks/facts-grounding,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Google DeepMind /models/gemini/flash/ table: Gemini 3 Flash Thinking, no tools by default."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 3 Flash, slug=gemini-3-flash-preview, provider=Google"
gemma-3-1b-it,facts_grounding,36.4,https://llm-stats.com/benchmarks/facts-grounding,"{""context"":""32k"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Google Gemma 3 1B instruction-tuned."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemma 3 1B, slug=gemma-3-1b-it, provider=Google"
grok-4.1,factscore,97.0,https://llm-stats.com/benchmarks/factscore,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per xAI Grok 4.1 blog: standard chart values. Hallucination/FActScore use with-web-search but those are not stored in BP."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Grok-4.1, slug=grok-4.1-2025-11-17, provider=xAI"
gemini-1.5-flash-8b,fleurs,86.4,https://llm-stats.com/benchmarks/fleurs,"{""context"":""1M"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Command A paper baseline."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 1.5 Flash 8B, slug=gemini-1.5-flash-8b, provider=Google"
gemini-1.5-flash,fleurs,9.6,https://llm-stats.com/benchmarks/fleurs,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemini 2.0 Flash Model Card (April 15 2025): non-thinking generation, default sampling. Same canonical setting applies to 1.5 / 2.0 family."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 1.5 Flash, slug=gemini-1.5-flash, provider=Google"
gemini-1.5-pro,fleurs,6.7,https://llm-stats.com/benchmarks/fleurs,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemini 2.0 Flash Model Card (April 15 2025): non-thinking generation, default sampling. Same canonical setting applies to 1.5 / 2.0 family."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 1.5 Pro, slug=gemini-1.5-pro, provider=Google"
qwen2.5-omni-7b,fleurs,4.1,https://llm-stats.com/benchmarks/fleurs,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5-Omni-7B, slug=qwen2.5-omni-7b, provider=Alibaba Cloud / Qwen Team"
o1-high,frontiermath,5.5,https://llm-stats.com/benchmarks/frontiermath,"{""effort"":""default (o1 base, not high)""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=o1, slug=o1-2024-12-17, provider=OpenAI"
gpt-5,frontiermath_tier4,26.3,https://llm-stats.com/benchmarks/frontiermath,"{""context"":""default"",""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per OpenAI GPT-5 dev blog. high reasoning effort."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-5, slug=gpt-5-2025-08-07, provider=OpenAI"
gpt-5-mini,frontiermath_tier4,22.1,https://llm-stats.com/benchmarks/frontiermath,"{""context"":""default"",""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per OpenAI GPT-5 developer blog https://openai.com/index/introducing-gpt-5-for-developers/"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-5 mini, slug=gpt-5-mini-2025-08-07, provider=OpenAI"
o3-high,frontiermath_tier4,15.8,https://llm-stats.com/benchmarks/frontiermath,"{""effort"":""default (o3 base, not high)""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=o3, slug=o3-2025-04-16, provider=OpenAI"
gpt-5-nano,frontiermath_tier4,9.6,https://llm-stats.com/benchmarks/frontiermath,"{""context"":""default"",""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per OpenAI GPT-5 developer blog https://openai.com/index/introducing-gpt-5-for-developers/"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-5 nano, slug=gpt-5-nano-2025-08-07, provider=OpenAI"
o1-high,frontiermath_tier4,5.5,https://llm-stats.com/benchmarks/frontiermath,"{""effort"":""default (o1 base, not high)""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=o1, slug=o1-2024-12-17, provider=OpenAI"
muse-spark,frontier_science_research,38.3,https://llm-stats.com/benchmarks/frontierscience-research,"{""effort"":""default"",""harness"":""internal (Meta MSL)"",""judge"":""rule-based or gpt-oss-120b (per benchmark)"",""mode"":""thinking"",""notes"":""Per Meta MSL blog https://ai.meta.com/blog/introducing-muse-spark-msl/. Standard Thinking mode. Contemplating mode results flagged for review."",""prompt_style"":""default"",""sampling"":""pass@1 (single-pass)"",""temperature"":""default"",""tools"":""none (per-benchmark override where tools are required)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Muse Spark, slug=muse-spark, provider=Meta"
gemini-diffusion,global_mmlu_lite,69.1,https://llm-stats.com/benchmarks/global-mmlu-lite,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini Diffusion, slug=gemini-diffusion, provider=Google"
gemma-3-1b-it,global_mmlu_lite,34.2,https://llm-stats.com/benchmarks/global-mmlu-lite,"{""context"":""32k"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Google Gemma 3 1B instruction-tuned."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemma 3 1B, slug=gemma-3-1b-it, provider=Google"
qwen3.6-plus,global_piqa,89.8,https://llm-stats.com/benchmarks/global-piqa,{},True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3.6 Plus, slug=qwen3.6-plus, provider=Alibaba Cloud / Qwen Team"
longcat-flash-chat,gpqa_diamond,73.2,https://llm-stats.com/benchmarks/gpqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=LongCat-Flash-Chat, slug=longcat-flash-chat, provider=Meituan"
ministral-8b,gpqa_diamond,66.8,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""n/a"",""harness"":""OLMES (Allen AI)"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OLMo 2 paper Table 7 Instruct (instruction-tuned variant)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Ministral 3 (8B Reasoning 2512), slug=ministral-8b-latest, provider=Mistral AI"
sarvam-30b,gpqa_diamond,66.5,https://llm-stats.com/benchmarks/gpqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Sarvam-30B, slug=sarvam-30b, provider=Sarvam AI"
grok-2,gpqa_diamond,56.0,https://llm-stats.com/benchmarks/gpqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Grok-2, slug=grok-2, provider=xAI"
grok-2-mini,gpqa_diamond,51.0,https://llm-stats.com/benchmarks/gpqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Grok-2 mini, slug=grok-2-mini, provider=xAI"
claude-3-opus,gpqa_diamond,50.4,https://llm-stats.com/benchmarks/gpqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3 Opus, slug=claude-3-opus-20240229, provider=Anthropic"
qwen-2.5-32b,gpqa_diamond,49.5,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""n/a"",""harness"":""OLMES (Allen AI)"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OLMo 2 paper Table 7 Instruct (instruction-tuned variant)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5 32B Instruct, slug=qwen-2.5-32b-instruct, provider=Alibaba Cloud / Qwen Team"
gpt-4-turbo,gpqa_diamond,48.0,https://llm-stats.com/benchmarks/gpqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-4 Turbo, slug=gpt-4-turbo-2024-04-09, provider=OpenAI"
mistral-small-24b,gpqa_diamond,45.3,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""n/a"",""harness"":""OLMES (Allen AI)"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OLMo 2 paper Table 7 Instruct (instruction-tuned variant)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Mistral Small 3 24B Instruct, slug=mistral-small-24b-instruct-2501, provider=Mistral AI"
qwen2-72b-instruct,gpqa_diamond,42.4,https://llm-stats.com/benchmarks/gpqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2 72B Instruct, slug=qwen2-72b-instruct, provider=Alibaba Cloud / Qwen Team"
nova-lite,gpqa_diamond,42.0,https://llm-stats.com/benchmarks/gpqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Nova Lite, slug=nova-lite, provider=Amazon"
llama-3.1-70b,gpqa_diamond,41.7,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""n/a"",""harness"":""OLMES (Allen AI)"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OLMo 2 paper Table 7 Instruct (instruction-tuned variant)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Llama 3.1 70B Instruct, slug=llama-3.1-70b-instruct, provider=Meta"
claude-3.5-haiku,gpqa_diamond,41.6,https://llm-stats.com/benchmarks/gpqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3.5 Haiku, slug=claude-3-5-haiku-20241022, provider=Anthropic"
claude-3-sonnet,gpqa_diamond,40.4,https://llm-stats.com/benchmarks/gpqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3 Sonnet, slug=claude-3-sonnet-20240229, provider=Anthropic"
gemini-diffusion,gpqa_diamond,40.4,https://llm-stats.com/benchmarks/gpqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini Diffusion, slug=gemini-diffusion, provider=Google"
nova-micro,gpqa_diamond,40.0,https://llm-stats.com/benchmarks/gpqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Nova Micro, slug=nova-micro, provider=Amazon"
gemini-1.5-flash-8b,gpqa_diamond,38.4,https://llm-stats.com/benchmarks/gpqa,"{""context"":""1M"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Command A paper baseline."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 1.5 Flash 8B, slug=gemini-1.5-flash-8b, provider=Google"
phi-3.5-moe-instruct,gpqa_diamond,36.8,https://llm-stats.com/benchmarks/gpqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Phi-3.5-MoE-instruct, slug=phi-3.5-moe-instruct, provider=Microsoft"
qwen-2.5-7b,gpqa_diamond,36.4,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""n/a"",""harness"":""OLMES (Allen AI)"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OLMo 2 paper Table 7 Instruct (instruction-tuned variant)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5 7B Instruct, slug=qwen-2.5-7b-instruct, provider=Alibaba Cloud / Qwen Team"
grok-1.5,gpqa_diamond,35.9,https://llm-stats.com/benchmarks/gpqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Grok-1.5, slug=grok-1.5, provider=xAI"
claude-3-haiku,gpqa_diamond,33.3,https://llm-stats.com/benchmarks/gpqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3 Haiku, slug=claude-3-haiku-20240307, provider=Anthropic"
gpt-3.5-turbo,gpqa_diamond,30.8,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""n/a"",""harness"":""OLMES (Allen AI)"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OLMo 2 paper Table 7 Instruct (instruction-tuned variant)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-3.5 Turbo, slug=gpt-3.5-turbo-0125, provider=OpenAI"
qwen2.5-omni-7b,gpqa_diamond,30.8,https://llm-stats.com/benchmarks/gpqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5-Omni-7B, slug=qwen2.5-omni-7b, provider=Alibaba Cloud / Qwen Team"
phi-3.5-mini-instruct,gpqa_diamond,30.4,https://llm-stats.com/benchmarks/gpqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Phi-3.5-mini-instruct, slug=phi-3.5-mini-instruct, provider=Microsoft"
qwen2-7b-instruct,gpqa_diamond,25.3,https://llm-stats.com/benchmarks/gpqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2 7B Instruct, slug=qwen2-7b-instruct, provider=Alibaba Cloud / Qwen Team"
claude-mythos,gpqa_main,94.6,https://llm-stats.com/benchmarks/gpqa,"{""context"":""up to 1M"",""effort"":""adaptive thinking at max effort"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Mythos System Card Table 6.3.A: standard config is adaptive thinking max effort, default sampling (temperature, top_p), avg 5 trials, context up to 1M."",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude Mythos Preview, slug=claude-mythos-preview, provider=Anthropic"
gemini-3.1-pro,gpqa_main,94.3,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Google DeepMind /models/gemini/pro/ table: Gemini 3.1 Pro Thinking (High), no tools by default."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 3.1 Pro, slug=gemini-3.1-pro-preview, provider=Google"
claude-opus-4.7,gpqa_main,94.2,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""default"",""harness"":""Terminus-2 (for terminal-bench, thinking disabled)"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Anthropic Opus 4.7 announcement table. Default thinking mode with extended thinking; Terminal-Bench uses Terminus-2 harness with thinking disabled (1\u00d7 guaranteed/3\u00d7 ceiling, avg 5 attempts)."",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude Opus 4.7, slug=claude-opus-4-7, provider=Anthropic"
gpt-5.5,gpqa_main,93.6,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per OpenAI gpt-5.5 blog: reasoning effort=xhigh, research environment (may differ from production ChatGPT)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-5.5, slug=gpt-5.5, provider=OpenAI"
gpt-5.2,gpqa_main,93.2,https://llm-stats.com/benchmarks/gpqa,"{""effort"":""pro""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-5.2 Pro, slug=gpt-5.2-pro-2025-12-11, provider=OpenAI"
gpt-5.4,gpqa_main,92.8,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per OpenAI gpt-5.4 blog: reasoning effort=xhigh, research environment (may differ from production ChatGPT)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-5.4, slug=gpt-5.4, provider=OpenAI"
gemini-3-pro,gpqa_main,91.9,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Google DeepMind /models/gemini/pro/ table: Gemini 3 Pro Thinking (High), no tools by default."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 3 Pro, slug=gemini-3-pro-preview, provider=Google"
claude-opus-4.6,gpqa_main,91.3,https://llm-stats.com/benchmarks/gpqa,"{""context"":""up to 1M"",""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Mythos System Card Table 6.3.A: standard config = adaptive thinking max effort, default sampling, avg 5 trials, context up to 1M."",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude Opus 4.6, slug=claude-opus-4-6, provider=Anthropic"
kimi-k2.6,gpqa_main,90.5,https://llm-stats.com/benchmarks/gpqa,"{""context"":""262144 (256K)"",""effort"":""max (98304 generation tokens)"",""harness"":""official (in-house SWE-agent-derived for SWE-Bench)"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Kimi K2.6 model card (HF moonshotai/Kimi-K2.6). Thinking mode, max effort, t=1.0 top_p=1.0."",""prompt_style"":""default"",""sampling"":""pass@1 (avg@10 coding, avg@3 vision)"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Kimi K2.6, slug=kimi-k2.6, provider=Moonshot AI"
gemini-3-flash,gpqa_main,90.4,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Google DeepMind /models/gemini/flash/ table: Gemini 3 Flash Thinking, no tools by default."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 3 Flash, slug=gemini-3-flash-preview, provider=Google"
qwen3.6-plus,gpqa_main,90.4,https://llm-stats.com/benchmarks/gpqa,{},True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3.6 Plus, slug=qwen3.6-plus, provider=Alibaba Cloud / Qwen Team"
deepseek-v4-pro,gpqa_main,90.1,https://llm-stats.com/benchmarks/gpqa,"{""effort"":""max""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek-V4-Pro-Max, slug=deepseek-v4-pro-max, provider=DeepSeek"
claude-sonnet-4.6,gpqa_main,89.9,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""max"",""harness"":""Terminus-2 (terminal-bench, thinking off); official otherwise"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Anthropic Sonnet 4.6 announcement: max thinking effort default; Terminal-Bench=thinking off, Terminus-2; SWE-bench=avg 10 trials; BrowseComp/HLE-with-tools have specific tool configs (web search/fetch + 50k context compaction)."",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5-15 trials)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude Sonnet 4.6, slug=claude-sonnet-4-6, provider=Anthropic"
muse-spark,gpqa_main,89.5,https://llm-stats.com/benchmarks/gpqa,"{""effort"":""default"",""harness"":""internal (Meta MSL)"",""judge"":""rule-based or gpt-oss-120b (per benchmark)"",""mode"":""thinking"",""notes"":""Per Meta MSL blog https://ai.meta.com/blog/introducing-muse-spark-msl/. Standard Thinking mode. Contemplating mode results flagged for review."",""prompt_style"":""default"",""sampling"":""pass@1 (single-pass)"",""temperature"":""default"",""tools"":""none (per-benchmark override where tools are required)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Muse Spark, slug=muse-spark, provider=Meta"
grok-4,gpqa_main,88.4,https://llm-stats.com/benchmarks/gpqa,"{""effort"":""heavy""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Grok-4 Heavy, slug=grok-4-heavy, provider=xAI"
gpt-5.1,gpqa_main,88.1,https://llm-stats.com/benchmarks/gpqa,"{""effort"":""high""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-5.1 High, slug=gpt-5.1-high-2025-11-12, provider=OpenAI"
deepseek-v4-flash,gpqa_main,88.1,https://llm-stats.com/benchmarks/gpqa,"{""effort"":""max""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek-V4-Flash-Max, slug=deepseek-v4-flash-max, provider=DeepSeek"
gpt-5,gpqa_main,88.1,https://llm-stats.com/benchmarks/gpqa,"{""effort"":""medium""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-5 Medium, slug=gpt-5-medium-2025-08-07, provider=OpenAI"
kimi-k2.5,gpqa_main,87.6,https://llm-stats.com/benchmarks/gpqa,"{""context"":""256K"",""effort"":""default (96k for reasoning, 64k for vision)"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Kimi K2.5 model card (HF moonshotai/Kimi-K2.5). Thinking mode, t=1.0 top_p=0.95."",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 AIME/HMMT, avg@8 GPQA, avg@3 vision, avg@4 Seal/WideSearch)"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Kimi K2.5, slug=kimi-k2.5, provider=Moonshot AI"
claude-opus-4.5,gpqa_main,87.0,https://llm-stats.com/benchmarks/gpqa,"{""context"":""200K (default)"",""effort"":""high (default)"",""harness"":""Terminus-2 (for terminal-bench); official otherwise"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Anthropic Opus 4.5 announcement: 64K thinking budget, interleaved scratchpads, 200K context, default effort=high, default sampling, avg 5 trials. Exceptions: SWE-bench Verified=no thinking; Terminal-Bench=128K thinking budget."",""prompt_style"":""interleaved scratchpads"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default (top_p, temperature)"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude Opus 4.5, slug=claude-opus-4-5-20251101, provider=Anthropic"
gemini-2.5-pro,gpqa_main,86.4,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""default"",""harness"":""AI Studio API"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Gemini 2.5 Pro Model Card (GA, June 27 2025): pass@1 single attempt, AI Studio API model-id gemini-2.5-pro-preview-06-05 / gemini-2.5-pro GA, default sampling. Release date set to GA date 2025-06-27."",""prompt_style"":""default"",""sampling"":""pass@1 (single attempt)"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 2.5 Pro Preview 06-05, slug=gemini-2.5-pro-preview-06-05, provider=Google"
glm-5.1,gpqa_main,86.2,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per GLM-5.1 blog."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GLM-5.1, slug=glm-5.1, provider=Zhipu AI"
glm-4.7,gpqa_main,85.7,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per GLM-5 blog."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GLM-4.7, slug=glm-4.7, provider=Zhipu AI"
claude-3.7-sonnet,gpqa_main,84.8,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Claude 4 blog cross-model table: with extended thinking for GPQA/MMMU/AIME/TAU; no extended thinking for SWE/Terminal/MMMLU."",""prompt_style"":""zeroshot-cot"",""sampling"":""pass@1"",""temperature"":0.0,""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3.7 Sonnet, slug=claude-3-7-sonnet-20250219, provider=Anthropic"
grok-3-beta,gpqa_main,84.6,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per xAI Grok 3 blog: canonical = non-reasoning mode (the cross-model table without 'Think' label). Reasoning ON values not used."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Grok-3, slug=grok-3, provider=xAI"
kimi-k2-thinking,gpqa_main,84.5,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default (256K)"",""effort"":""default (96k thinking-token budget for HLE/AIME/HMMT/GPQA; 128k for IMO/LCB/OJ)"",""harness"":""official"",""judge"":""rule-based (HLE uses o3-mini)"",""mode"":""thinking"",""notes"":""Per Kimi K2 Thinking blog (moonshotai.github.io/Kimi-K2/thinking.html). Thinking model."",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 for AIME/HMMT no-tools)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Kimi K2-Thinking-0905, slug=kimi-k2-thinking-0905, provider=Moonshot AI"
claude-sonnet-4.5,gpqa_main,83.4,https://llm-stats.com/benchmarks/gpqa,"{""context"":""200K (default)"",""effort"":""high (default)"",""harness"":""Terminus-2 (for terminal-bench); official otherwise"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Anthropic Opus 4.5 announcement: 64K thinking budget, interleaved scratchpads, 200K context, default effort=high, default sampling, avg 5 trials. Exceptions: SWE-bench Verified=no thinking; Terminal-Bench=128K thinking budget."",""prompt_style"":""interleaved scratchpads"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default (top_p, temperature)"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude Sonnet 4.5, slug=claude-sonnet-4-5-20250929, provider=Anthropic"
o3-high,gpqa_main,83.3,https://llm-stats.com/benchmarks/gpqa,"{""effort"":""default (o3 base, not high)""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=o3, slug=o3-2025-04-16, provider=OpenAI"
gemini-2.5-flash,gpqa_main,82.8,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default (1M)"",""effort"":""default"",""harness"":""AI Studio API (gemini-2.5-flash GA)"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Gemini 2.5 Flash Model Card (Dec 2025) GA Thinking column."",""prompt_style"":""default"",""sampling"":""pass@1 (single attempt; multi-trial avg for smaller bench)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 2.5 Flash, slug=gemini-2.5-flash, provider=Google"
deepseek-v3.2,gpqa_main,82.4,https://llm-stats.com/benchmarks/gpqa,"{""mode"":""thinking (DeepSeek-Reasoner endpoint)""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek-V3.2 (Thinking), slug=deepseek-reasoner, provider=DeepSeek"
gpt-5-mini,gpqa_main,82.3,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per OpenAI GPT-5 developer blog https://openai.com/index/introducing-gpt-5-for-developers/"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-5 mini, slug=gpt-5-mini-2025-08-07, provider=OpenAI"
o4-mini-high,gpqa_main,81.4,https://llm-stats.com/benchmarks/gpqa,"{""effort"":""default (o4-mini base, not high)""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=o4-mini, slug=o4-mini, provider=OpenAI"
qwen3-235b,gpqa_main,81.1,https://llm-stats.com/benchmarks/gpqa,"{""mode"":""thinking"",""snapshot"":""2507""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3-235B-A22B-Thinking-2507, slug=qwen3-235b-a22b-thinking-2507, provider=Alibaba Cloud / Qwen Team"
glm-4.6,gpqa_main,81.0,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per GLM-4.7 blog."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GLM-4.6, slug=glm-4.6, provider=Zhipu AI"
deepseek-r1,gpqa_main,81.0,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per DeepSeek R1-0528 model card."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek-R1-0528, slug=deepseek-r1-0528, provider=DeepSeek"
minimax-m2.1,gpqa_main,81.0,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per MiniMax M2.5 model card (referenced as predecessor)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=MiniMax M2.1, slug=minimax-m2.1, provider=MiniMax"
claude-opus-4.1,gpqa_main,80.9,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""default"",""harness"":""Terminus 1 (for agentic benchmarks)"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Anthropic Opus 4.1 announcement table. Single-pass values; harness=Terminus 1 for Terminal-Bench (note: differs from Claude 4 blog which used Claude Code framework)."",""prompt_style"":""default"",""sampling"":""pass@1 (single-pass)"",""temperature"":""top_p=0.95"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude Opus 4.1, slug=claude-opus-4-1-20250805, provider=Anthropic"
gpt-oss-120b,gpqa_main,80.9,https://llm-stats.com/benchmarks/gpqa,"{""effort"":""high""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT OSS 120B High, slug=gpt-oss-120b-high, provider=OpenAI"
claude-opus-4,gpqa_main,79.6,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""default"",""harness"":""Claude Code agent framework (for agentic benchmarks)"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Anthropic Claude 4 announcement table: single-pass values; no extended thinking for SWE/Terminal/MMMLU; extended thinking up to 64K for GPQA/MMMU/AIME/TAU; tools=agentic for SWE/Terminal/TAU/OSWorld. Agentic harness=Claude Code per Claude 4 blog."",""prompt_style"":""default"",""sampling"":""pass@1 (single-pass)"",""temperature"":""top_p=0.95"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude Opus 4, slug=claude-opus-4-20250514, provider=Anthropic"
minimax-m2,gpqa_main,78.0,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per MiniMax M2 model card."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=MiniMax M2, slug=minimax-m2, provider=MiniMax"
o1-high,gpqa_main,78.0,https://llm-stats.com/benchmarks/gpqa,"{""effort"":""default (o1 base, not high)""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=o1, slug=o1-2024-12-17, provider=OpenAI"
nemotron-ultra-253b,gpqa_main,76.0,https://llm-stats.com/benchmarks/gpqa,"{""context"":""32k"",""effort"":""default"",""harness"":""NVIDIA eval pipeline (transformers/vLLM)"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per NVIDIA HF model card https://huggingface.co/nvidia/Llama-3_1-Nemotron-Ultra-253B-v1: Reasoning On mode, temperature=0.6, top_p=0.95, 32k context, benchmarks run up to 16x and averaged."",""prompt_style"":""custom prompt templates per benchmark"",""sampling"":""pass@1 (avg of up to 16 benchmark runs)"",""temperature"":""0.6 (thinking mode)"",""tools"":""none (benchmark-specified)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Llama 3.1 Nemotron Ultra 253B v1, slug=llama-3.1-nemotron-ultra-253b-v1, provider=NVIDIA"
kimi-k2,gpqa_main,75.8,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default (256K)"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Kimi K2 tech report (arxiv:2507.20534) Table 3 Kimi-K2-Instruct column. Non-thinking model."",""prompt_style"":""default"",""sampling"":""pass@1 (avg of 4-64 trials per benchmark)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Kimi K2 0905, slug=kimi-k2-0905, provider=Moonshot AI"
claude-sonnet-4,gpqa_main,75.4,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""default"",""harness"":""Claude Code agent framework (for agentic benchmarks)"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Anthropic Claude 4 announcement table: single-pass values; no extended thinking for SWE/Terminal/MMMLU; extended thinking up to 64K for GPQA/MMMU/AIME/TAU; tools=agentic for SWE/Terminal/TAU/OSWorld. Agentic harness=Claude Code per Claude 4 blog."",""prompt_style"":""default"",""sampling"":""pass@1 (single-pass)"",""temperature"":""top_p=0.95"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude Sonnet 4, slug=claude-sonnet-4-20250514, provider=Anthropic"
gpt-oss-20b,gpqa_main,74.2,https://llm-stats.com/benchmarks/gpqa,"{""effort"":""high""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT OSS 20B High, slug=gpt-oss-20b-high, provider=OpenAI"
gemini-2.0-flash,gpqa_main,74.2,https://llm-stats.com/benchmarks/gpqa,"{""effort"":""thinking""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 2.0 Flash Thinking, slug=gemini-2.0-flash-thinking, provider=Google"
longcat-flash-chat,gpqa_main,73.2,https://llm-stats.com/benchmarks/gpqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=LongCat-Flash-Chat, slug=longcat-flash-chat, provider=Meituan"
claude-haiku-4.5,gpqa_main,73.0,https://llm-stats.com/benchmarks/gpqa,"{""effort"":""128K thinking budget"",""harness"":""official"",""judge"":""official harness"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1 (avg over \u226510 runs unless noted)"",""temperature"":""default (Anthropic default)"",""tools"":""none (per-bench override for agentic benchmarks)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude Haiku 4.5, slug=claude-haiku-4-5-20251001, provider=Anthropic"
gpt-5-nano,gpqa_main,71.2,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per OpenAI GPT-5 developer blog https://openai.com/index/introducing-gpt-5-for-developers/"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-5 nano, slug=gpt-5-nano-2025-08-07, provider=OpenAI"
gpt-4o,gpqa_main,70.1,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OpenAI GPT-4.1 blog table."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-4o, slug=gpt-4o-2024-08-06, provider=OpenAI"
llama-4-maverick,gpqa_main,69.8,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""n/a"",""harness"":""official-meta"",""judge"":""rule-based"",""mode"":""non-thinking"",""multimodal_input"":true,""prompt_style"":""zeroshot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Llama 4 Maverick, slug=llama-4-maverick, provider=Meta"
gpt-4.5,gpqa_main,69.5,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OpenAI GPT-4.1 blog table."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-4.5, slug=gpt-4.5, provider=OpenAI"
phi-4-reasoning-plus,gpqa_main,68.9,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default (32K)"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Phi-4-reasoning paper. Same as phi-4-reasoning + GRPO RL on top."",""prompt_style"":""default"",""sampling"":""pass@1 (avg of 5)"",""temperature"":""0.8"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Phi 4 Reasoning Plus, slug=phi-4-reasoning-plus, provider=Microsoft"
sarvam-30b,gpqa_main,66.5,https://llm-stats.com/benchmarks/gpqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Sarvam-30B, slug=sarvam-30b, provider=Sarvam AI"
gpt-4.1,gpqa_main,66.3,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OpenAI GPT-5 dev blog. non-reasoning model."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-4.1, slug=gpt-4.1-2025-04-14, provider=OpenAI"
phi-4-reasoning,gpqa_main,65.8,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default (32K)"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Phi-4-reasoning paper (arxiv:2504.21318) Tables 1-2. Thinking model, temp 0.8, max_seq_len 31k."",""prompt_style"":""default"",""sampling"":""pass@1 (avg of 5)"",""temperature"":""0.8"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Phi 4 Reasoning, slug=phi-4-reasoning, provider=Microsoft"
qwen3-30b-a3b,gpqa_main,65.8,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""default"",""harness"":""official (Qwen-Agent / sglang)"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Qwen3 tech report (arxiv:2505.09388) Table 13/15/17/19 Thinking column."",""prompt_style"":""default (enable_thinking=True)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3 30B A3B, slug=qwen3-30b-a3b, provider=Alibaba Cloud / Qwen Team"
deepseek-r1-distill-llama-70b,gpqa_main,65.2,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""default"",""harness"":""official (DeepSeek SGLang/vLLM)"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Distill of DeepSeek-R1 onto Llama-70B base. Cells in BP currently primarily sourced from Qwen3 tech report Table 13 third-party self-test (matches_canonical=false). DeepSeek own paper audit pending \u2014 canonical_setting may need refinement."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek R1 Distill Llama 70B, slug=deepseek-r1-distill-llama-70b, provider=DeepSeek"
qwq-32b,gpqa_main,65.2,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default (32K)"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per QwQ-32B release blog (qwenlm.github.io/blog/qwq-32b/) + Qwen3 tech report Table 13 baseline column."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=QwQ-32B, slug=qwq-32b, provider=Alibaba Cloud / Qwen Team"
gpt-4.1-mini,gpqa_main,65.0,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OpenAI GPT-5 dev blog. non-reasoning model."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-4.1 mini, slug=gpt-4.1-mini-2025-04-14, provider=OpenAI"
deepseek-r1-distill-qwen-32b,gpqa_main,62.1,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per DeepSeek R1 paper Table 15 (distilled model)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek R1 Distill Qwen 32B, slug=deepseek-r1-distill-qwen-32b, provider=DeepSeek"
o1-mini,gpqa_main,60.0,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Phi-4-reasoning paper Table 1 baseline (Microsoft self-test)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=o1-mini, slug=o1-mini, provider=OpenAI"
deepseek-r1-distill-qwen-14b,gpqa_main,59.1,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per DeepSeek R1 paper Table 15 (distilled model)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek R1 Distill Qwen 14B, slug=deepseek-r1-distill-qwen-14b, provider=DeepSeek"
gemini-1.5-pro,gpqa_main,59.1,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemini 2.0 Flash Model Card (April 15 2025): non-thinking generation, default sampling. Same canonical setting applies to 1.5 / 2.0 family."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 1.5 Pro, slug=gemini-1.5-pro, provider=Google"
llama-4-scout,gpqa_main,57.2,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""n/a"",""harness"":""official-meta"",""judge"":""rule-based"",""mode"":""non-thinking"",""multimodal_input"":true,""prompt_style"":""zeroshot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Llama 4 Scout, slug=llama-4-scout, provider=Meta"
phi-4,gpqa_main,56.1,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default (16K)"",""effort"":""n/a"",""harness"":""OpenAI simple-evals"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Phi-4 paper (arxiv:2412.08905) Table 1, simple-evals framework, temperature 0.5."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.5"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Phi 4, slug=phi-4, provider=Microsoft"
grok-2,gpqa_main,56.0,https://llm-stats.com/benchmarks/gpqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Grok-2, slug=grok-2, provider=xAI"
gemini-2.0-flash-lite,gpqa_main,51.5,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemini 2.0 Flash Model Card (April 15 2025): non-thinking generation, default sampling. Same canonical setting applies to 1.5 / 2.0 family."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 2.0 Flash-Lite, slug=gemini-2.0-flash-lite, provider=Google"
grok-2-mini,gpqa_main,51.0,https://llm-stats.com/benchmarks/gpqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Grok-2 mini, slug=grok-2-mini, provider=xAI"
gemini-1.5-flash,gpqa_main,51.0,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemini 2.0 Flash Model Card (April 15 2025): non-thinking generation, default sampling. Same canonical setting applies to 1.5 / 2.0 family."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 1.5 Flash, slug=gemini-1.5-flash, provider=Google"
claude-3-opus,gpqa_main,50.4,https://llm-stats.com/benchmarks/gpqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3 Opus, slug=claude-3-opus-20240229, provider=Anthropic"
gpt-4.1-nano,gpqa_main,50.3,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OpenAI GPT-5 dev blog. non-reasoning model."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-4.1 nano, slug=gpt-4.1-nano-2025-04-14, provider=OpenAI"
qwen-2.5-32b,gpqa_main,49.5,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""n/a"",""harness"":""OLMES (Allen AI)"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OLMo 2 paper Table 7 Instruct (instruction-tuned variant)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5 32B Instruct, slug=qwen-2.5-32b-instruct, provider=Alibaba Cloud / Qwen Team"
deepseek-r1-distill-qwen-7b,gpqa_main,49.1,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per DeepSeek R1 paper Table 15 (distilled model)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek R1 Distill Qwen 7B, slug=deepseek-r1-distill-qwen-7b, provider=DeepSeek"
deepseek-r1-distill-llama-8b,gpqa_main,49.0,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per DeepSeek R1 paper Table 15 (distilled model)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek R1 Distill Llama 8B, slug=deepseek-r1-distill-llama-8b, provider=DeepSeek"
gpt-4-turbo,gpqa_main,48.0,https://llm-stats.com/benchmarks/gpqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-4 Turbo, slug=gpt-4-turbo-2024-04-09, provider=OpenAI"
mistral-small-3.1,gpqa_main,46.0,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""n/a"",""harness"":""mistral-eval"",""judge"":""rule-based"",""mode"":""non-thinking"",""multimodal_input"":true,""prompt_style"":""mixed (5-shot CoT for MMLU/Pro/GPQA per HF MC; default for others)"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Mistral Small 3.1 24B Instruct, slug=mistral-small-3.1-24b-instruct-2503, provider=Mistral AI"
qwen-2.5-14b-instruct,gpqa_main,45.5,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default (128K)"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Phi-4 paper Table 1 baseline."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5 14B Instruct, slug=qwen-2.5-14b-instruct, provider=Alibaba Cloud / Qwen Team"
mistral-small-24b,gpqa_main,45.3,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""n/a"",""harness"":""OLMES (Allen AI)"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OLMo 2 paper Table 7 Instruct (instruction-tuned variant)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Mistral Small 3 24B Instruct, slug=mistral-small-24b-instruct-2501, provider=Mistral AI"
qwen2-72b-instruct,gpqa_main,42.4,https://llm-stats.com/benchmarks/gpqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2 72B Instruct, slug=qwen2-72b-instruct, provider=Alibaba Cloud / Qwen Team"
gemma-3-27b,gpqa_main,42.4,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default (128K)"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemma 3 tech report (arxiv:2503.19786) Table 6 IT column for 27B."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemma 3 27B, slug=gemma-3-27b-it, provider=Google"
nova-lite,gpqa_main,42.0,https://llm-stats.com/benchmarks/gpqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Nova Lite, slug=nova-lite, provider=Amazon"
llama-3.1-70b,gpqa_main,41.7,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""n/a"",""harness"":""OLMES (Allen AI)"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OLMo 2 paper Table 7 Instruct (instruction-tuned variant)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Llama 3.1 70B Instruct, slug=llama-3.1-70b-instruct, provider=Meta"
claude-3.5-haiku,gpqa_main,41.6,https://llm-stats.com/benchmarks/gpqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3.5 Haiku, slug=claude-3-5-haiku-20241022, provider=Anthropic"
gemma-3-12b,gpqa_main,40.9,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemma 3 tech report Table 6 IT instruction-tuned column."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemma 3 12B, slug=gemma-3-12b-it, provider=Google"
claude-3-sonnet,gpqa_main,40.4,https://llm-stats.com/benchmarks/gpqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3 Sonnet, slug=claude-3-sonnet-20240229, provider=Anthropic"
gemini-diffusion,gpqa_main,40.4,https://llm-stats.com/benchmarks/gpqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini Diffusion, slug=gemini-diffusion, provider=Google"
gpt-4o-mini,gpqa_main,40.2,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OpenAI GPT-4.1 blog table."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-4o mini, slug=gpt-4o-mini-2024-07-18, provider=OpenAI"
nova-micro,gpqa_main,40.0,https://llm-stats.com/benchmarks/gpqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Nova Micro, slug=nova-micro, provider=Amazon"
phi-3.5-moe-instruct,gpqa_main,36.8,https://llm-stats.com/benchmarks/gpqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Phi-3.5-MoE-instruct, slug=phi-3.5-moe-instruct, provider=Microsoft"
qwen-2.5-7b,gpqa_main,36.4,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""n/a"",""harness"":""OLMES (Allen AI)"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OLMo 2 paper Table 7 Instruct (instruction-tuned variant)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5 7B Instruct, slug=qwen-2.5-7b-instruct, provider=Alibaba Cloud / Qwen Team"
grok-1.5,gpqa_main,35.9,https://llm-stats.com/benchmarks/gpqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Grok-1.5, slug=grok-1.5, provider=xAI"
deepseek-r1-distill-qwen-1.5b,gpqa_main,33.8,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per DeepSeek R1 paper Table 15 (distilled model)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek R1 Distill Qwen 1.5B, slug=deepseek-r1-distill-qwen-1.5b, provider=DeepSeek"
claude-3-haiku,gpqa_main,33.3,https://llm-stats.com/benchmarks/gpqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3 Haiku, slug=claude-3-haiku-20240307, provider=Anthropic"
gpt-3.5-turbo,gpqa_main,30.8,https://llm-stats.com/benchmarks/gpqa,"{""context"":""default"",""effort"":""n/a"",""harness"":""OLMES (Allen AI)"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OLMo 2 paper Table 7 Instruct (instruction-tuned variant)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-3.5 Turbo, slug=gpt-3.5-turbo-0125, provider=OpenAI"
qwen2.5-omni-7b,gpqa_main,30.8,https://llm-stats.com/benchmarks/gpqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5-Omni-7B, slug=qwen2.5-omni-7b, provider=Alibaba Cloud / Qwen Team"
phi-3.5-mini-instruct,gpqa_main,30.4,https://llm-stats.com/benchmarks/gpqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Phi-3.5-mini-instruct, slug=phi-3.5-mini-instruct, provider=Microsoft"
qwen2-7b-instruct,gpqa_main,25.3,https://llm-stats.com/benchmarks/gpqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2 7B Instruct, slug=qwen2-7b-instruct, provider=Alibaba Cloud / Qwen Team"
gemma-3-1b-it,gpqa_main,19.2,https://llm-stats.com/benchmarks/gpqa,"{""context"":""32k"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Google Gemma 3 1B instruction-tuned."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemma 3 1B, slug=gemma-3-1b-it, provider=Google"
claude-mythos,graphwalks_bfs_128k_plus,80.0,https://llm-stats.com/benchmarks/graphwalks-bfs->128k,"{""context"":""up to 1M"",""effort"":""adaptive thinking at max effort"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Mythos System Card Table 6.3.A: standard config is adaptive thinking max effort, default sampling (temperature, top_p), avg 5 trials, context up to 1M."",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude Mythos Preview, slug=claude-mythos-preview, provider=Anthropic"
claude-opus-4.6,graphwalks_bfs_128k_plus,61.5,https://llm-stats.com/benchmarks/graphwalks-bfs->128k,"{""context"":""up to 1M"",""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Mythos System Card Table 6.3.A: standard config = adaptive thinking max effort, default sampling, avg 5 trials, context up to 1M."",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude Opus 4.6, slug=claude-opus-4-6, provider=Anthropic"
gpt-5.5,graphwalks_bfs_128k_plus,45.4,https://llm-stats.com/benchmarks/graphwalks-bfs->128k,"{""context"":""default"",""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per OpenAI gpt-5.5 blog: reasoning effort=xhigh, research environment (may differ from production ChatGPT)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-5.5, slug=gpt-5.5, provider=OpenAI"
gpt-5.4,graphwalks_bfs_128k_plus,21.4,https://llm-stats.com/benchmarks/graphwalks-bfs->128k,"{""context"":""default"",""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per OpenAI gpt-5.4 blog: reasoning effort=xhigh, research environment (may differ from production ChatGPT)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-5.4, slug=gpt-5.4, provider=OpenAI"
claude-opus-4.6,graphwalks_parents_128k_plus,95.4,https://llm-stats.com/benchmarks/graphwalks-parents->128k,"{""context"":""up to 1M"",""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Mythos System Card Table 6.3.A: standard config = adaptive thinking max effort, default sampling, avg 5 trials, context up to 1M."",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude Opus 4.6, slug=claude-opus-4-6, provider=Anthropic"
gpt-5.5,graphwalks_parents_128k_plus,58.5,https://llm-stats.com/benchmarks/graphwalks-parents->128k,"{""context"":""default"",""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per OpenAI gpt-5.5 blog: reasoning effort=xhigh, research environment (may differ from production ChatGPT)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-5.5, slug=gpt-5.5, provider=OpenAI"
gpt-5.4,graphwalks_parents_128k_plus,32.4,https://llm-stats.com/benchmarks/graphwalks-parents->128k,"{""context"":""default"",""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per OpenAI gpt-5.4 blog: reasoning effort=xhigh, research environment (may differ from production ChatGPT)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-5.4, slug=gpt-5.4, provider=OpenAI"
kimi-k2,gsm8k,97.3,https://llm-stats.com/benchmarks/gsm8k,"{""variant"":""instruct (BP kimi-k2 is base)""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Kimi K2 Instruct, slug=kimi-k2-instruct, provider=Moonshot AI"
o1-high,gsm8k,97.1,https://llm-stats.com/benchmarks/gsm8k,"{""effort"":""default (o1 base, not high)""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=o1, slug=o1-2024-12-17, provider=OpenAI"
gpt-4.5,gsm8k,97.0,https://llm-stats.com/benchmarks/gsm8k,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OpenAI GPT-4.1 blog table."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-4.5, slug=gpt-4.5, provider=OpenAI"
llama-3.1-405b-instruct,gsm8k,96.8,https://llm-stats.com/benchmarks/gsm8k,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per DeepSeek V3 paper Table 6."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Llama 3.1 405B Instruct, slug=llama-3.1-405b-instruct, provider=Meta"
claude-3.5-sonnet,gsm8k,96.4,https://llm-stats.com/benchmarks/gsm8k,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per DeepSeek V3 paper Table 6."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3.5 Sonnet, slug=claude-3-5-sonnet-20241022, provider=Anthropic"
deepseek-v2.5-0905,gsm8k,95.1,https://llm-stats.com/benchmarks/gsm8k,"{""snapshot"":""0905""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek-V2.5, slug=deepseek-v2.5, provider=DeepSeek"
claude-3-opus,gsm8k,95.0,https://llm-stats.com/benchmarks/gsm8k,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3 Opus, slug=claude-3-opus-20240229, provider=Anthropic"
qwen-2.5-14b-instruct,gsm8k,94.8,https://llm-stats.com/benchmarks/gsm8k,"{""context"":""default (128K)"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Phi-4 paper Table 1 baseline."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5 14B Instruct, slug=qwen-2.5-14b-instruct, provider=Alibaba Cloud / Qwen Team"
nova-lite,gsm8k,94.5,https://llm-stats.com/benchmarks/gsm8k,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Nova Lite, slug=nova-lite, provider=Amazon"
gemma-3-12b,gsm8k,94.4,https://llm-stats.com/benchmarks/gsm8k,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemma 3 tech report Table 6 IT instruction-tuned column."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemma 3 12B, slug=gemma-3-12b-it, provider=Google"
mistral-large-2,gsm8k,93.0,https://llm-stats.com/benchmarks/gsm8k,"{""context"":""128K"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Command A paper (third-party Cohere reproduction baseline)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Mistral Large 2, slug=mistral-large-2-2407, provider=Mistral AI"
claude-3-sonnet,gsm8k,92.3,https://llm-stats.com/benchmarks/gsm8k,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3 Sonnet, slug=claude-3-sonnet-20240229, provider=Anthropic"
nova-micro,gsm8k,92.3,https://llm-stats.com/benchmarks/gsm8k,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Nova Micro, slug=nova-micro, provider=Amazon"
qwen2-72b-instruct,gsm8k,91.1,https://llm-stats.com/benchmarks/gsm8k,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2 72B Instruct, slug=qwen2-72b-instruct, provider=Alibaba Cloud / Qwen Team"
gemini-1.5-pro,gsm8k,90.8,https://llm-stats.com/benchmarks/gsm8k,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemini 2.0 Flash Model Card (April 15 2025): non-thinking generation, default sampling. Same canonical setting applies to 1.5 / 2.0 family."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 1.5 Pro, slug=gemini-1.5-pro, provider=Google"
grok-1.5,gsm8k,90.0,https://llm-stats.com/benchmarks/gsm8k,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Grok-1.5, slug=grok-1.5, provider=xAI"
claude-3-haiku,gsm8k,88.9,https://llm-stats.com/benchmarks/gsm8k,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3 Haiku, slug=claude-3-haiku-20240307, provider=Anthropic"
qwen2.5-omni-7b,gsm8k,88.7,https://llm-stats.com/benchmarks/gsm8k,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5-Omni-7B, slug=qwen2.5-omni-7b, provider=Alibaba Cloud / Qwen Team"
phi-3.5-moe-instruct,gsm8k,88.7,https://llm-stats.com/benchmarks/gsm8k,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Phi-3.5-MoE-instruct, slug=phi-3.5-moe-instruct, provider=Microsoft"
gemini-1.5-flash,gsm8k,86.2,https://llm-stats.com/benchmarks/gsm8k,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemini 2.0 Flash Model Card (April 15 2025): non-thinking generation, default sampling. Same canonical setting applies to 1.5 / 2.0 family."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 1.5 Flash, slug=gemini-1.5-flash, provider=Google"
phi-3.5-mini-instruct,gsm8k,86.2,https://llm-stats.com/benchmarks/gsm8k,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Phi-3.5-mini-instruct, slug=phi-3.5-mini-instruct, provider=Microsoft"
qwen2-7b-instruct,gsm8k,82.3,https://llm-stats.com/benchmarks/gsm8k,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2 7B Instruct, slug=qwen2-7b-instruct, provider=Alibaba Cloud / Qwen Team"
granite-3.3-8b-instruct,gsm8k,80.9,https://llm-stats.com/benchmarks/gsm8k,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Granite 3.3 8B Instruct, slug=granite-3.3-8b-instruct, provider=IBM"
granite-3.3-8b-base,gsm8k,59.0,https://llm-stats.com/benchmarks/gsm8k,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Granite 3.3 8B Base, slug=granite-3.3-8b-base, provider=IBM"
gpt-5,healthbench_hard,1.6,https://llm-stats.com/benchmarks/healthbench-hard,"{""context"":""default"",""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per OpenAI GPT-5 dev blog. high reasoning effort."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-5, slug=gpt-5-2025-08-07, provider=OpenAI"
claude-3-opus,hellaswag,95.4,https://llm-stats.com/benchmarks/hellaswag,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3 Opus, slug=claude-3-opus-20240229, provider=Anthropic"
gemini-1.5-pro,hellaswag,93.3,https://llm-stats.com/benchmarks/hellaswag,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemini 2.0 Flash Model Card (April 15 2025): non-thinking generation, default sampling. Same canonical setting applies to 1.5 / 2.0 family."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 1.5 Pro, slug=gemini-1.5-pro, provider=Google"
claude-3-sonnet,hellaswag,89.0,https://llm-stats.com/benchmarks/hellaswag,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3 Sonnet, slug=claude-3-sonnet-20240229, provider=Anthropic"
qwen2-72b-instruct,hellaswag,87.6,https://llm-stats.com/benchmarks/hellaswag,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2 72B Instruct, slug=qwen2-72b-instruct, provider=Alibaba Cloud / Qwen Team"
gemini-1.5-flash,hellaswag,86.5,https://llm-stats.com/benchmarks/hellaswag,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemini 2.0 Flash Model Card (April 15 2025): non-thinking generation, default sampling. Same canonical setting applies to 1.5 / 2.0 family."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 1.5 Flash, slug=gemini-1.5-flash, provider=Google"
claude-3-haiku,hellaswag,85.9,https://llm-stats.com/benchmarks/hellaswag,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3 Haiku, slug=claude-3-haiku-20240307, provider=Anthropic"
qwen-2.5-32b,hellaswag,85.2,https://llm-stats.com/benchmarks/hellaswag,"{""context"":""default"",""effort"":""n/a"",""harness"":""OLMES (Allen AI)"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OLMo 2 paper Table 7 Instruct (instruction-tuned variant)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5 32B Instruct, slug=qwen-2.5-32b-instruct, provider=Alibaba Cloud / Qwen Team"
phi-3.5-moe-instruct,hellaswag,83.8,https://llm-stats.com/benchmarks/hellaswag,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Phi-3.5-MoE-instruct, slug=phi-3.5-moe-instruct, provider=Microsoft"
granite-3.3-8b-base,hellaswag,80.1,https://llm-stats.com/benchmarks/hellaswag,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Granite 3.3 8B Base, slug=granite-3.3-8b-base, provider=IBM"
phi-3.5-mini-instruct,hellaswag,69.4,https://llm-stats.com/benchmarks/hellaswag,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Phi-3.5-mini-instruct, slug=phi-3.5-mini-instruct, provider=Microsoft"
phi-4-mini,hellaswag,69.1,https://llm-stats.com/benchmarks/hellaswag,"{""context"":""default (128K)"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Phi-4-mini-instruct HF model card."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Phi 4 Mini, slug=phi-4-mini, provider=Microsoft"
gemini-1.5-flash-8b,hiddenmath,32.8,https://llm-stats.com/benchmarks/hiddenmath,"{""context"":""1M"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Command A paper baseline."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 1.5 Flash 8B, slug=gemini-1.5-flash-8b, provider=Google"
gemma-3-1b-it,hiddenmath,15.8,https://llm-stats.com/benchmarks/hiddenmath,"{""context"":""32k"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Google Gemma 3 1B instruction-tuned."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemma 3 1B, slug=gemma-3-1b-it, provider=Google"
qwen3.6-plus,hmmt_feb_2025,96.7,https://llm-stats.com/benchmarks/hmmt-2025,{},True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3.6 Plus, slug=qwen3.6-plus, provider=Alibaba Cloud / Qwen Team"
glm-5.1,hmmt_feb_2025,94.0,https://llm-stats.com/benchmarks/hmmt-2025,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per GLM-5.1 blog."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GLM-5.1, slug=glm-5.1, provider=Zhipu AI"
sarvam-30b,hmmt_feb_2025,73.3,https://llm-stats.com/benchmarks/hmmt-2025,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Sarvam-30B, slug=sarvam-30b, provider=Sarvam AI"
granite-3.3-8b-base,humaneval_plus,86.1,https://llm-stats.com/benchmarks/humaneval+,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Granite 3.3 8B Base, slug=granite-3.3-8b-base, provider=IBM"
granite-3.3-8b-instruct,humaneval_plus,86.1,https://llm-stats.com/benchmarks/humaneval+,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Granite 3.3 8B Instruct, slug=granite-3.3-8b-instruct, provider=IBM"
qwen-2.5-32b,humaneval_plus,52.4,https://llm-stats.com/benchmarks/humaneval+,"{""context"":""default"",""effort"":""n/a"",""harness"":""OLMES (Allen AI)"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OLMo 2 paper Table 7 Instruct (instruction-tuned variant)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5 32B Instruct, slug=qwen-2.5-32b-instruct, provider=Alibaba Cloud / Qwen Team"
o1-high,humaneval,88.1,https://llm-stats.com/benchmarks/humaneval,"{""effort"":""default (o1 base, not high)""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=o1, slug=o1-2024-12-17, provider=OpenAI"
mistral-small-24b,humaneval,84.8,https://llm-stats.com/benchmarks/humaneval,"{""context"":""default"",""effort"":""n/a"",""harness"":""OLMES (Allen AI)"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OLMo 2 paper Table 7 Instruct (instruction-tuned variant)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Mistral Small 3 24B Instruct, slug=mistral-small-24b-instruct-2501, provider=Mistral AI"
gemma-3-1b-it,humaneval,41.5,https://llm-stats.com/benchmarks/humaneval,"{""context"":""32k"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Google Gemma 3 1B instruction-tuned."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemma 3 1B, slug=gemma-3-1b-it, provider=Google"
gpt-4o,hle,5.3,https://llm-stats.com/benchmarks/humanity's-last-exam,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OpenAI GPT-4.1 blog table."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-4o, slug=gpt-4o-2024-08-06, provider=OpenAI"
claude-mythos,hle_text,64.7,https://llm-stats.com/benchmarks/humanity's-last-exam,"{""context"":""up to 1M"",""effort"":""adaptive thinking at max effort"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Mythos System Card Table 6.3.A: standard config is adaptive thinking max effort, default sampling (temperature, top_p), avg 5 trials, context up to 1M."",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude Mythos Preview, slug=claude-mythos-preview, provider=Anthropic"
muse-spark,hle_text,58.4,https://llm-stats.com/benchmarks/humanity's-last-exam,"{""effort"":""default"",""harness"":""internal (Meta MSL)"",""judge"":""rule-based or gpt-oss-120b (per benchmark)"",""mode"":""thinking"",""notes"":""Per Meta MSL blog https://ai.meta.com/blog/introducing-muse-spark-msl/. Standard Thinking mode. Contemplating mode results flagged for review."",""prompt_style"":""default"",""sampling"":""pass@1 (single-pass)"",""temperature"":""default"",""tools"":""none (per-benchmark override where tools are required)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Muse Spark, slug=muse-spark, provider=Meta"
gpt-5.5,hle_text,46.9,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,"Hy3 official model-card appendix: HLE (no tools, text-only) = 46.9*. Tencent own testing."
claude-opus-4.7,hle_text,54.7,https://llm-stats.com/benchmarks/humanity's-last-exam,"{""context"":""default"",""effort"":""default"",""harness"":""Terminus-2 (for terminal-bench, thinking disabled)"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Anthropic Opus 4.7 announcement table. Default thinking mode with extended thinking; Terminal-Bench uses Terminus-2 harness with thinking disabled (1\u00d7 guaranteed/3\u00d7 ceiling, avg 5 attempts)."",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude Opus 4.7, slug=claude-opus-4-7, provider=Anthropic"
claude-opus-4.6,hle_text,53.1,https://llm-stats.com/benchmarks/humanity's-last-exam,"{""context"":""up to 1M"",""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Mythos System Card Table 6.3.A: standard config = adaptive thinking max effort, default sampling, avg 5 trials, context up to 1M."",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude Opus 4.6, slug=claude-opus-4-6, provider=Anthropic"
glm-5.1,hle_text,31.0,https://z.ai/blog/glm-5.2,"{""context"":""default"",""effort"":""official reported"",""harness"":""official reasoning evaluation"",""judge"":""GPT-5.5 medium for AIME/HMMT/IMO; benchmark-specified otherwise"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none"",""top_p"":""0.95""}",True,official_blog,verified,GLM-5.2 official release source: GLM-5.1 / HLE = 31.0.
gemini-3.1-pro,hle_text,45.0,https://z.ai/blog/glm-5.2,"{""context"":""default"",""effort"":""high"",""harness"":""official reasoning evaluation"",""judge"":""GPT-5.5 medium for AIME/HMMT/IMO; benchmark-specified otherwise"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none"",""top_p"":""0.95""}",True,official_blog,verified,GLM-5.2 official release source: Gemini 3.1 Pro / HLE = 45.0.
kimi-k2-thinking,hle_text,51.0,https://llm-stats.com/benchmarks/humanity's-last-exam,"{""context"":""default (256K)"",""effort"":""default (96k thinking-token budget for HLE/AIME/HMMT/GPQA; 128k for IMO/LCB/OJ)"",""harness"":""official"",""judge"":""rule-based (HLE uses o3-mini)"",""mode"":""thinking"",""notes"":""Per Kimi K2 Thinking blog (moonshotai.github.io/Kimi-K2/thinking.html). Thinking model."",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 for AIME/HMMT no-tools)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Kimi K2-Thinking-0905, slug=kimi-k2-thinking-0905, provider=Moonshot AI"
grok-4,hle_text,50.7,https://llm-stats.com/benchmarks/humanity's-last-exam,"{""effort"":""heavy""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Grok-4 Heavy, slug=grok-4-heavy, provider=xAI"
kimi-k2.5,hle_text,50.2,https://llm-stats.com/benchmarks/humanity's-last-exam,"{""context"":""256K"",""effort"":""default (96k for reasoning, 64k for vision)"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Kimi K2.5 model card (HF moonshotai/Kimi-K2.5). Thinking mode, t=1.0 top_p=0.95."",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 AIME/HMMT, avg@8 GPQA, avg@3 vision, avg@4 Seal/WideSearch)"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Kimi K2.5, slug=kimi-k2.5, provider=Moonshot AI"
claude-sonnet-4.6,hle_text,49.0,https://llm-stats.com/benchmarks/humanity's-last-exam,"{""context"":""default"",""effort"":""max"",""harness"":""Terminus-2 (terminal-bench, thinking off); official otherwise"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Anthropic Sonnet 4.6 announcement: max thinking effort default; Terminal-Bench=thinking off, Terminus-2; SWE-bench=avg 10 trials; BrowseComp/HLE-with-tools have specific tool configs (web search/fetch + 50k context compaction)."",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5-15 trials)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude Sonnet 4.6, slug=claude-sonnet-4-6, provider=Anthropic"
deepseek-v4-pro,hle_text,37.7,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""context"":""default"",""effort"":""Max"",""harness"":""official reasoning evaluation"",""judge"":""GPT-5.5 medium for AIME/HMMT/IMO; benchmark-specified otherwise"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none"",""top_p"":""0.95""}",True,model_card,verified,DeepSeek-V4-Pro first-party source: HLE = 37.7; GLM-5.2 table used as cross-check.
deepseek-v4-flash,hle_text,34.8,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,"Hy3 official model-card appendix: HLE (no tools, text-only) = 34.8. Comparator-reported value."
gemini-3-flash,hle_text,43.5,https://llm-stats.com/benchmarks/humanity's-last-exam,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Google DeepMind /models/gemini/flash/ table: Gemini 3 Flash Thinking, no tools by default."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 3 Flash, slug=gemini-3-flash-preview, provider=Google"
glm-4.7,hle_text,42.8,https://llm-stats.com/benchmarks/humanity's-last-exam,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per GLM-5 blog."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GLM-4.7, slug=glm-4.7, provider=Zhipu AI"
deepseek-v3.2,hle_text,40.8,https://llm-stats.com/benchmarks/humanity's-last-exam,"{""context"":""128K"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per DeepSeek V3.2 tech report: temperature=1.0, context=128K, thinking mode for tool-use."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek-V3.2, slug=deepseek-v3.2, provider=DeepSeek"
gpt-5.4,hle_text,39.8,https://llm-stats.com/benchmarks/humanity's-last-exam,"{""context"":""default"",""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per OpenAI gpt-5.4 blog: reasoning effort=xhigh, research environment (may differ from production ChatGPT)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-5.4, slug=gpt-5.4, provider=OpenAI"
kimi-k2.6,hle_text,36.4,https://llm-stats.com/benchmarks/humanity's-last-exam,"{""context"":""262144 (256K)"",""effort"":""max (98304 generation tokens)"",""harness"":""official (in-house SWE-agent-derived for SWE-Bench)"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Kimi K2.6 model card (HF moonshotai/Kimi-K2.6). Thinking mode, max effort, t=1.0 top_p=1.0."",""prompt_style"":""default"",""sampling"":""pass@1 (avg@10 coding, avg@3 vision)"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Kimi K2.6, slug=kimi-k2.6, provider=Moonshot AI"
deepseek-v3.2-speciale,hle_text,30.6,https://llm-stats.com/benchmarks/humanity's-last-exam,"{""context"":""128K"",""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""DS V3.2-Speciale: high-compute reasoning variant of V3.2. Per V3.2 tech report Table 3."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek-V3.2-Speciale, slug=deepseek-v3.2-speciale, provider=DeepSeek"
qwen3.6-plus,hle_text,28.8,https://llm-stats.com/benchmarks/humanity's-last-exam,{},True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3.6 Plus, slug=qwen3.6-plus, provider=Alibaba Cloud / Qwen Team"
gpt-5,hle_text,24.8,https://llm-stats.com/benchmarks/humanity's-last-exam,"{""context"":""default"",""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per OpenAI GPT-5 dev blog. high reasoning effort."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-5, slug=gpt-5-2025-08-07, provider=OpenAI"
minimax-m2.1,hle_text,22.0,https://llm-stats.com/benchmarks/humanity's-last-exam,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per MiniMax M2.5 model card (referenced as predecessor)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=MiniMax M2.1, slug=minimax-m2.1, provider=MiniMax"
gemini-2.5-pro,hle_text,21.6,https://llm-stats.com/benchmarks/humanity's-last-exam,"{""context"":""default"",""effort"":""default"",""harness"":""AI Studio API"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Gemini 2.5 Pro Model Card (GA, June 27 2025): pass@1 single attempt, AI Studio API model-id gemini-2.5-pro-preview-06-05 / gemini-2.5-pro GA, default sampling. Release date set to GA date 2025-06-27."",""prompt_style"":""default"",""sampling"":""pass@1 (single attempt)"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 2.5 Pro Preview 06-05, slug=gemini-2.5-pro-preview-06-05, provider=Google"
qwen3-235b,hle_text,18.2,https://llm-stats.com/benchmarks/humanity's-last-exam,"{""mode"":""thinking"",""snapshot"":""2507""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3-235B-A22B-Thinking-2507, slug=qwen3-235b-a22b-thinking-2507, provider=Alibaba Cloud / Qwen Team"
deepseek-r1,hle_text,17.7,https://llm-stats.com/benchmarks/humanity's-last-exam,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per DeepSeek R1-0528 model card."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek-R1-0528, slug=deepseek-r1-0528, provider=DeepSeek"
glm-4.6,hle_text,17.2,https://llm-stats.com/benchmarks/humanity's-last-exam,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per GLM-4.7 blog."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GLM-4.6, slug=glm-4.6, provider=Zhipu AI"
gpt-5-mini,hle_text,16.7,https://llm-stats.com/benchmarks/humanity's-last-exam,"{""context"":""default"",""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per OpenAI GPT-5 developer blog https://openai.com/index/introducing-gpt-5-for-developers/"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-5 mini, slug=gpt-5-mini-2025-08-07, provider=OpenAI"
gpt-oss-120b,hle_text,14.9,https://llm-stats.com/benchmarks/humanity's-last-exam,"{""context"":""default (128K)"",""effort"":""high"",""harness"":""official (harmony chat format)"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per OpenAI gpt-oss model card (arxiv:2508.10925) Table 3, high reasoning effort. Default with-tools for agentic; AIME/GPQA also has no-tools variant in same table."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""with tools (per-bench override; harmony chat format default)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT OSS 120B, slug=gpt-oss-120b, provider=OpenAI"
o4-mini-high,hle_text,14.7,https://llm-stats.com/benchmarks/humanity's-last-exam,"{""effort"":""default (o4-mini base, not high)""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=o4-mini, slug=o4-mini, provider=OpenAI"
o3-high,hle_text,14.7,https://llm-stats.com/benchmarks/humanity's-last-exam,"{""effort"":""default (o3 base, not high)""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=o3, slug=o3-2025-04-16, provider=OpenAI"
minimax-m2,hle_text,12.5,https://llm-stats.com/benchmarks/humanity's-last-exam,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per MiniMax M2 model card."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=MiniMax M2, slug=minimax-m2, provider=MiniMax"
gemini-2.5-flash,hle_text,11.0,https://llm-stats.com/benchmarks/humanity's-last-exam,"{""context"":""default (1M)"",""effort"":""default"",""harness"":""AI Studio API (gemini-2.5-flash GA)"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Gemini 2.5 Flash Model Card (Dec 2025) GA Thinking column."",""prompt_style"":""default"",""sampling"":""pass@1 (single attempt; multi-trial avg for smaller bench)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 2.5 Flash, slug=gemini-2.5-flash, provider=Google"
gpt-oss-20b,hle_text,10.9,https://llm-stats.com/benchmarks/humanity's-last-exam,"{""context"":""default (128K)"",""effort"":""high"",""harness"":""official (harmony chat format)"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per OpenAI gpt-oss model card (arxiv:2508.10925) Table 3, high reasoning effort. Default with-tools for agentic; AIME/GPQA also has no-tools variant in same table."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""with tools (per-bench override; harmony chat format default)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT OSS 20B, slug=gpt-oss-20b, provider=OpenAI"
gpt-5-nano,hle_text,8.7,https://llm-stats.com/benchmarks/humanity's-last-exam,"{""context"":""default"",""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per OpenAI GPT-5 developer blog https://openai.com/index/introducing-gpt-5-for-developers/"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-5 nano, slug=gpt-5-nano-2025-08-07, provider=OpenAI"
gpt-4.1,hle_text,5.4,https://llm-stats.com/benchmarks/humanity's-last-exam,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OpenAI GPT-5 dev blog. non-reasoning model."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-4.1, slug=gpt-4.1-2025-04-14, provider=OpenAI"
gpt-4o,hle_text,5.3,https://llm-stats.com/benchmarks/humanity's-last-exam,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OpenAI GPT-4.1 blog table."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-4o, slug=gpt-4o-2024-08-06, provider=OpenAI"
kimi-k2,hle_text,4.7,https://llm-stats.com/benchmarks/humanity's-last-exam,"{""snapshot"":""0905"",""variant"":""instruct""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Kimi K2-Instruct-0905, slug=kimi-k2-instruct-0905, provider=Moonshot AI"
gpt-4.1-mini,hle_text,3.7,https://llm-stats.com/benchmarks/humanity's-last-exam,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OpenAI GPT-5 dev blog. non-reasoning model."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-4.1 mini, slug=gpt-4.1-mini-2025-04-14, provider=OpenAI"
gpt-oss-120b,ifbench,69.5,https://llm-stats.com/benchmarks/ifbench,"{""effort"":""high""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT OSS 120B High, slug=gpt-oss-120b-high, provider=OpenAI"
qwen3.6-plus,ifeval,94.3,https://llm-stats.com/benchmarks/ifeval,{},True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3.6 Plus, slug=qwen3.6-plus, provider=Alibaba Cloud / Qwen Team"
nova-lite,ifeval,89.7,https://llm-stats.com/benchmarks/ifeval,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Nova Lite, slug=nova-lite, provider=Amazon"
longcat-flash-chat,ifeval,89.6,https://llm-stats.com/benchmarks/ifeval,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=LongCat-Flash-Chat, slug=longcat-flash-chat, provider=Meituan"
gemma-3-12b,ifeval,88.9,https://llm-stats.com/benchmarks/ifeval,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemma 3 tech report Table 6 IT instruction-tuned column."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemma 3 12B, slug=gemma-3-12b-it, provider=Google"
nova-micro,ifeval,87.2,https://llm-stats.com/benchmarks/ifeval,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Nova Micro, slug=nova-micro, provider=Amazon"
minicpm-sala,ifeval,76.3,https://llm-stats.com/benchmarks/ifeval,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=MiniCPM-SALA, slug=minicpm-sala, provider=OpenBMB"
granite-3.3-8b-instruct,ifeval,74.8,https://llm-stats.com/benchmarks/ifeval,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Granite 3.3 8B Instruct, slug=granite-3.3-8b-instruct, provider=IBM"
granite-3.3-8b-base,ifeval,74.8,https://llm-stats.com/benchmarks/ifeval,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Granite 3.3 8B Base, slug=granite-3.3-8b-base, provider=IBM"
pixtral-12b,ifeval,61.3,https://llm-stats.com/benchmarks/ifeval,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Pixtral-12B, slug=pixtral-12b-2409, provider=Mistral AI"
gemma-3-27b,infovqa,70.6,https://llm-stats.com/benchmarks/infovqa,"{""context"":""default (128K)"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemma 3 tech report (arxiv:2503.19786) Table 6 IT column for 27B."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemma 3 27B, slug=gemma-3-27b-it, provider=Google"
gemma-3-12b,infovqa,64.9,https://llm-stats.com/benchmarks/infovqa,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemma 3 tech report Table 6 IT instruction-tuned column."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemma 3 12B, slug=gemma-3-12b-it, provider=Google"
o1-high,livebench,67.0,https://llm-stats.com/benchmarks/livebench,"{""effort"":""default (o1 base, not high)""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=o1, slug=o1-2024-12-17, provider=OpenAI"
qwen-2.5-7b,livebench,35.9,https://llm-stats.com/benchmarks/livebench,"{""context"":""default"",""effort"":""n/a"",""harness"":""OLMES (Allen AI)"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OLMo 2 paper Table 7 Instruct (instruction-tuned variant)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5 7B Instruct, slug=qwen-2.5-7b-instruct, provider=Alibaba Cloud / Qwen Team"
qwen2.5-omni-7b,livebench,29.6,https://llm-stats.com/benchmarks/livebench,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5-Omni-7B, slug=qwen2.5-omni-7b, provider=Alibaba Cloud / Qwen Team"
muse-spark,livecodebench_pro,80.0,https://llm-stats.com/benchmarks/livecodebench-pro,"{""effort"":""default"",""harness"":""internal (Meta MSL)"",""judge"":""rule-based or gpt-oss-120b (per benchmark)"",""mode"":""thinking"",""notes"":""Per Meta MSL blog https://ai.meta.com/blog/introducing-muse-spark-msl/. Standard Thinking mode. Contemplating mode results flagged for review."",""prompt_style"":""default"",""sampling"":""pass@1 (single-pass)"",""temperature"":""default"",""tools"":""none (per-benchmark override where tools are required)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Muse Spark, slug=muse-spark, provider=Meta"
gemini-2.5-flash,livecodebench_v5,63.9,https://llm-stats.com/benchmarks/livecodebench-v5,"{""context"":""default (1M)"",""effort"":""default"",""harness"":""AI Studio API (gemini-2.5-flash GA)"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Gemini 2.5 Flash Model Card (Dec 2025) GA Thinking column."",""prompt_style"":""default"",""sampling"":""pass@1 (single attempt; multi-trial avg for smaller bench)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 2.5 Flash, slug=gemini-2.5-flash, provider=Google"
minicpm-sala,livecodebench_v5,60.5,https://llm-stats.com/benchmarks/livecodebench-v5,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=MiniCPM-SALA, slug=minicpm-sala, provider=OpenBMB"
gemini-2.0-flash-lite,livecodebench_v5,28.9,https://llm-stats.com/benchmarks/livecodebench-v5,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemini 2.0 Flash Model Card (April 15 2025): non-thinking generation, default sampling. Same canonical setting applies to 1.5 / 2.0 family."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 2.0 Flash-Lite, slug=gemini-2.0-flash-lite, provider=Google"
qwen3.6-plus,livecodebench_v6,87.1,https://llm-stats.com/benchmarks/livecodebench-v6,{},True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3.6 Plus, slug=qwen3.6-plus, provider=Alibaba Cloud / Qwen Team"
gpt-oss-120b,livecodebench_v6,81.9,https://llm-stats.com/benchmarks/livecodebench-v6,"{""effort"":""high""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT OSS 120B High, slug=gpt-oss-120b-high, provider=OpenAI"
qwen3-235b,livecodebench_v6,74.1,https://llm-stats.com/benchmarks/livecodebench-v6,"{""mode"":""thinking"",""snapshot"":""2507""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3-235B-A22B-Thinking-2507, slug=qwen3-235b-a22b-thinking-2507, provider=Alibaba Cloud / Qwen Team"
sarvam-30b,livecodebench_v6,70.0,https://llm-stats.com/benchmarks/livecodebench-v6,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Sarvam-30B, slug=sarvam-30b, provider=Sarvam AI"
minicpm-sala,livecodebench_v6,52.0,https://llm-stats.com/benchmarks/livecodebench-v6,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=MiniCPM-SALA, slug=minicpm-sala, provider=OpenBMB"
minimax-m2.1,livecodebench,78.0,https://llm-stats.com/benchmarks/livecodebench,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per MiniMax M2.5 model card (referenced as predecessor)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=MiniMax M2.1, slug=minimax-m2.1, provider=MiniMax"
ministral-8b,livecodebench,61.6,https://llm-stats.com/benchmarks/livecodebench,"{""context"":""default"",""effort"":""n/a"",""harness"":""OLMES (Allen AI)"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OLMo 2 paper Table 7 Instruct (instruction-tuned variant)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Ministral 3 (8B Reasoning 2512), slug=ministral-8b-latest, provider=Mistral AI"
longcat-flash-chat,livecodebench,48.0,https://llm-stats.com/benchmarks/livecodebench,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=LongCat-Flash-Chat, slug=longcat-flash-chat, provider=Meituan"
gemini-diffusion,livecodebench,30.9,https://llm-stats.com/benchmarks/livecodebench,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini Diffusion, slug=gemini-diffusion, provider=Google"
qwen-2.5-7b,livecodebench,28.7,https://llm-stats.com/benchmarks/livecodebench,"{""context"":""default"",""effort"":""n/a"",""harness"":""OLMES (Allen AI)"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OLMo 2 paper Table 7 Instruct (instruction-tuned variant)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5 7B Instruct, slug=qwen-2.5-7b-instruct, provider=Alibaba Cloud / Qwen Team"
qwen2-7b-instruct,livecodebench,26.6,https://llm-stats.com/benchmarks/livecodebench,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2 7B Instruct, slug=qwen2-7b-instruct, provider=Alibaba Cloud / Qwen Team"
gemma-3-1b-it,livecodebench,1.9,https://llm-stats.com/benchmarks/livecodebench,"{""context"":""32k"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Google Gemma 3 1B instruction-tuned."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemma 3 1B, slug=gemma-3-1b-it, provider=Google"
qwen3.6-plus,longbench_v2,62.0,https://llm-stats.com/benchmarks/longbench-v2,{},True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3.6 Plus, slug=qwen3.6-plus, provider=Alibaba Cloud / Qwen Team"
nova-lite,lvbench,40.4,https://llm-stats.com/benchmarks/lvbench,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Nova Lite, slug=nova-lite, provider=Amazon"
sarvam-30b,math_500,97.0,https://llm-stats.com/benchmarks/math-500,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Sarvam-30B, slug=sarvam-30b, provider=Sarvam AI"
longcat-flash-chat,math_500,96.4,https://llm-stats.com/benchmarks/math-500,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=LongCat-Flash-Chat, slug=longcat-flash-chat, provider=Meituan"
o1-mini,math_500,90.0,https://llm-stats.com/benchmarks/math-500,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Phi-4-reasoning paper Table 1 baseline (Microsoft self-test)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=o1-mini, slug=o1-mini, provider=OpenAI"
granite-3.3-8b-instruct,math_500,69.0,https://llm-stats.com/benchmarks/math-500,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Granite 3.3 8B Instruct, slug=granite-3.3-8b-instruct, provider=IBM"
granite-3.3-8b-base,math_500,69.0,https://llm-stats.com/benchmarks/math-500,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Granite 3.3 8B Base, slug=granite-3.3-8b-base, provider=IBM"
o1-high,math,96.4,https://llm-stats.com/benchmarks/math,"{""effort"":""default (o1 base, not high)""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=o1, slug=o1-2024-12-17, provider=OpenAI"
mistral-large-3,math,90.4,https://llm-stats.com/benchmarks/math,{},True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Mistral Large 3, slug=mistral-large-3-2509, provider=Mistral AI"
kimi-k2,math,89.1,https://llm-stats.com/benchmarks/math,"{""context"":""default (256K)"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Kimi K2 tech report (arxiv:2507.20534) Table 3 Kimi-K2-Instruct column. Non-thinking model."",""prompt_style"":""default"",""sampling"":""pass@1 (avg of 4-64 trials per benchmark)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Kimi K2 0905, slug=kimi-k2-0905, provider=Moonshot AI"
gpt-5,math,84.7,https://llm-stats.com/benchmarks/math,"{""context"":""default"",""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per OpenAI GPT-5 dev blog. high reasoning effort."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-5, slug=gpt-5-2025-08-07, provider=OpenAI"
phi-4,math,80.4,https://llm-stats.com/benchmarks/math,"{""context"":""default (16K)"",""effort"":""n/a"",""harness"":""OpenAI simple-evals"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Phi-4 paper (arxiv:2412.08905) Table 1, simple-evals framework, temperature 0.5."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.5"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Phi 4, slug=phi-4, provider=Microsoft"
qwen-2.5-14b-instruct,math,80.0,https://llm-stats.com/benchmarks/math,"{""context"":""default (128K)"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Phi-4 paper Table 1 baseline."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5 14B Instruct, slug=qwen-2.5-14b-instruct, provider=Alibaba Cloud / Qwen Team"
claude-3.5-sonnet,math,78.3,https://llm-stats.com/benchmarks/math,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per DeepSeek V3 paper Table 6."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3.5 Sonnet, slug=claude-3-5-sonnet-20241022, provider=Anthropic"
gpt-4o,math,76.6,https://llm-stats.com/benchmarks/math,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OpenAI GPT-4.1 blog table."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-4o, slug=gpt-4o-2024-05-13, provider=OpenAI"
grok-2,math,76.1,https://llm-stats.com/benchmarks/math,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Grok-2, slug=grok-2, provider=xAI"
deepseek-v2.5-0905,math,74.7,https://llm-stats.com/benchmarks/math,"{""snapshot"":""0905""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek-V2.5, slug=deepseek-v2.5, provider=DeepSeek"
llama-3.1-405b-instruct,math,73.8,https://llm-stats.com/benchmarks/math,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per DeepSeek V3 paper Table 6."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Llama 3.1 405B Instruct, slug=llama-3.1-405b-instruct, provider=Meta"
nova-lite,math,73.3,https://llm-stats.com/benchmarks/math,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Nova Lite, slug=nova-lite, provider=Amazon"
grok-2-mini,math,73.0,https://llm-stats.com/benchmarks/math,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Grok-2 mini, slug=grok-2-mini, provider=xAI"
gpt-4-turbo,math,72.6,https://llm-stats.com/benchmarks/math,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-4 Turbo, slug=gpt-4-turbo-2024-04-09, provider=OpenAI"
qwen3-235b,math,71.8,https://llm-stats.com/benchmarks/math,"{""context"":""default"",""effort"":""default"",""harness"":""official (Qwen-Agent / sglang)"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Qwen3 tech report (arxiv:2505.09388) Table 11 Thinking column. Output length 32768 (38912 for AIME)."",""prompt_style"":""default (enable_thinking=True)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3 235B A22B, slug=qwen3-235b-a22b, provider=Alibaba Cloud / Qwen Team"
qwen2.5-omni-7b,math,71.5,https://llm-stats.com/benchmarks/math,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5-Omni-7B, slug=qwen2.5-omni-7b, provider=Alibaba Cloud / Qwen Team"
claude-3.5-haiku,math,69.4,https://llm-stats.com/benchmarks/math,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3.5 Haiku, slug=claude-3-5-haiku-20241022, provider=Anthropic"
nova-micro,math,69.3,https://llm-stats.com/benchmarks/math,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Nova Micro, slug=nova-micro, provider=Amazon"
llama-4-maverick,math,61.2,https://llm-stats.com/benchmarks/math,"{""context"":""default"",""effort"":""n/a"",""harness"":""official-meta"",""judge"":""rule-based"",""mode"":""non-thinking"",""multimodal_input"":true,""prompt_style"":""zeroshot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Llama 4 Maverick, slug=llama-4-maverick, provider=Meta"
claude-3-opus,math,60.1,https://llm-stats.com/benchmarks/math,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3 Opus, slug=claude-3-opus-20240229, provider=Anthropic"
qwen2-72b-instruct,math,59.7,https://llm-stats.com/benchmarks/math,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2 72B Instruct, slug=qwen2-72b-instruct, provider=Alibaba Cloud / Qwen Team"
phi-3.5-moe-instruct,math,59.5,https://llm-stats.com/benchmarks/math,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Phi-3.5-MoE-instruct, slug=phi-3.5-moe-instruct, provider=Microsoft"
gemini-1.5-flash-8b,math,58.7,https://llm-stats.com/benchmarks/math,"{""context"":""1M"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Command A paper baseline."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 1.5 Flash 8B, slug=gemini-1.5-flash-8b, provider=Google"
grok-1.5,math,50.6,https://llm-stats.com/benchmarks/math,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Grok-1.5, slug=grok-1.5, provider=xAI"
llama-4-scout,math,50.3,https://llm-stats.com/benchmarks/math,"{""context"":""default"",""effort"":""n/a"",""harness"":""official-meta"",""judge"":""rule-based"",""mode"":""non-thinking"",""multimodal_input"":true,""prompt_style"":""zeroshot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Llama 4 Scout, slug=llama-4-scout, provider=Meta"
qwen2-7b-instruct,math,49.6,https://llm-stats.com/benchmarks/math,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2 7B Instruct, slug=qwen2-7b-instruct, provider=Alibaba Cloud / Qwen Team"
phi-3.5-mini-instruct,math,48.5,https://llm-stats.com/benchmarks/math,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Phi-3.5-mini-instruct, slug=phi-3.5-mini-instruct, provider=Microsoft"
pixtral-12b,math,48.1,https://llm-stats.com/benchmarks/math,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Pixtral-12B, slug=pixtral-12b-2409, provider=Mistral AI"
gemma-3-1b-it,math,48.0,https://llm-stats.com/benchmarks/math,"{""context"":""32k"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Google Gemma 3 1B instruction-tuned."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemma 3 1B, slug=gemma-3-1b-it, provider=Google"
claude-3-sonnet,math,43.1,https://llm-stats.com/benchmarks/math,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3 Sonnet, slug=claude-3-sonnet-20240229, provider=Anthropic"
claude-3-haiku,math,38.9,https://llm-stats.com/benchmarks/math,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3 Haiku, slug=claude-3-haiku-20240307, provider=Anthropic"
qwen3.6-plus,mathvision,88.0,https://llm-stats.com/benchmarks/mathvision,{},True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3.6 Plus, slug=qwen3.6-plus, provider=Alibaba Cloud / Qwen Team"
qwen2.5-omni-7b,mathvision,25.0,https://llm-stats.com/benchmarks/mathvision,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5-Omni-7B, slug=qwen2.5-omni-7b, provider=Alibaba Cloud / Qwen Team"
gemma-3-27b,mathvista_mini,67.6,https://llm-stats.com/benchmarks/mathvista-mini,"{""context"":""default (128K)"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemma 3 tech report (arxiv:2503.19786) Table 6 IT column for 27B."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemma 3 27B, slug=gemma-3-27b-it, provider=Google"
gemma-3-12b,mathvista_mini,62.9,https://llm-stats.com/benchmarks/mathvista-mini,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemma 3 tech report Table 6 IT instruction-tuned column."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemma 3 12B, slug=gemma-3-12b-it, provider=Google"
o3-high,mathvista,86.8,https://llm-stats.com/benchmarks/mathvista,"{""effort"":""default (o3 base, not high)""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=o3, slug=o3-2025-04-16, provider=OpenAI"
o4-mini-high,mathvista,84.3,https://llm-stats.com/benchmarks/mathvista,"{""effort"":""default (o4-mini base, not high)""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=o4-mini, slug=o4-mini, provider=OpenAI"
llama-4-maverick,mathvista,73.7,https://llm-stats.com/benchmarks/mathvista,"{""context"":""default"",""effort"":""n/a"",""harness"":""official-meta"",""judge"":""rule-based"",""mode"":""non-thinking"",""multimodal_input"":true,""prompt_style"":""zeroshot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Llama 4 Maverick, slug=llama-4-maverick, provider=Meta"
llama-4-scout,mathvista,70.7,https://llm-stats.com/benchmarks/mathvista,"{""context"":""default"",""effort"":""n/a"",""harness"":""official-meta"",""judge"":""rule-based"",""mode"":""non-thinking"",""multimodal_input"":true,""prompt_style"":""zeroshot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Llama 4 Scout, slug=llama-4-scout, provider=Meta"
grok-2,mathvista,69.0,https://llm-stats.com/benchmarks/mathvista,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Grok-2, slug=grok-2, provider=xAI"
grok-2-mini,mathvista,68.1,https://llm-stats.com/benchmarks/mathvista,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Grok-2 mini, slug=grok-2-mini, provider=xAI"
gemini-1.5-pro,mathvista,68.1,https://llm-stats.com/benchmarks/mathvista,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemini 2.0 Flash Model Card (April 15 2025): non-thinking generation, default sampling. Same canonical setting applies to 1.5 / 2.0 family."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 1.5 Pro, slug=gemini-1.5-pro, provider=Google"
qwen2.5-omni-7b,mathvista,67.9,https://llm-stats.com/benchmarks/mathvista,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5-Omni-7B, slug=qwen2.5-omni-7b, provider=Alibaba Cloud / Qwen Team"
claude-3.5-sonnet,mathvista,67.7,https://llm-stats.com/benchmarks/mathvista,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per DeepSeek V3 paper Table 6."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3.5 Sonnet, slug=claude-3-5-sonnet-20241022, provider=Anthropic"
gemini-1.5-flash,mathvista,65.8,https://llm-stats.com/benchmarks/mathvista,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemini 2.0 Flash Model Card (April 15 2025): non-thinking generation, default sampling. Same canonical setting applies to 1.5 / 2.0 family."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 1.5 Flash, slug=gemini-1.5-flash, provider=Google"
pixtral-12b,mathvista,58.0,https://llm-stats.com/benchmarks/mathvista,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Pixtral-12B, slug=pixtral-12b-2409, provider=Mistral AI"
gemini-1.5-flash-8b,mathvista,54.7,https://llm-stats.com/benchmarks/mathvista,"{""context"":""1M"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Command A paper baseline."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 1.5 Flash 8B, slug=gemini-1.5-flash-8b, provider=Google"
grok-1.5,mathvista,52.8,https://llm-stats.com/benchmarks/mathvista,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Grok-1.5, slug=grok-1.5, provider=xAI"
gpt-3.5-turbo,mathvista,0.0,https://llm-stats.com/benchmarks/mathvista,"{""context"":""default"",""effort"":""n/a"",""harness"":""OLMES (Allen AI)"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OLMo 2 paper Table 7 Instruct (instruction-tuned variant)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-3.5 Turbo, slug=gpt-3.5-turbo-0125, provider=OpenAI"
qwen-2.5-32b,mbpp_plus,67.2,https://llm-stats.com/benchmarks/mbpp+,"{""context"":""default"",""effort"":""n/a"",""harness"":""OLMES (Allen AI)"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OLMo 2 paper Table 7 Instruct (instruction-tuned variant)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5 32B Instruct, slug=qwen-2.5-32b-instruct, provider=Alibaba Cloud / Qwen Team"
qwen-2.5-14b-instruct,mbpp_plus,63.2,https://llm-stats.com/benchmarks/mbpp+,"{""context"":""default (128K)"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Phi-4 paper Table 1 baseline."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5 14B Instruct, slug=qwen-2.5-14b-instruct, provider=Alibaba Cloud / Qwen Team"
qwen3.6-plus,mcpmark,48.2,https://llm-stats.com/benchmarks/mcp-mark,{},True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3.6 Plus, slug=qwen3.6-plus, provider=Alibaba Cloud / Qwen Team"
llama-4-maverick,mgsm,92.3,https://llm-stats.com/benchmarks/mgsm,"{""context"":""default"",""effort"":""n/a"",""harness"":""official-meta"",""judge"":""rule-based"",""mode"":""non-thinking"",""multimodal_input"":true,""prompt_style"":""zeroshot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Llama 4 Maverick, slug=llama-4-maverick, provider=Meta"
claude-3.5-sonnet,mgsm,91.6,https://llm-stats.com/benchmarks/mgsm,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per DeepSeek V3 paper Table 6."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3.5 Sonnet, slug=claude-3-5-sonnet-20240620, provider=Anthropic"
claude-3-opus,mgsm,90.7,https://llm-stats.com/benchmarks/mgsm,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3 Opus, slug=claude-3-opus-20240229, provider=Anthropic"
llama-4-scout,mgsm,90.6,https://llm-stats.com/benchmarks/mgsm,"{""context"":""default"",""effort"":""n/a"",""harness"":""official-meta"",""judge"":""rule-based"",""mode"":""non-thinking"",""multimodal_input"":true,""prompt_style"":""zeroshot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Llama 4 Scout, slug=llama-4-scout, provider=Meta"
gpt-4o,mgsm,90.5,https://llm-stats.com/benchmarks/mgsm,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OpenAI GPT-4.1 blog table."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-4o, slug=gpt-4o-2024-05-13, provider=OpenAI"
o1-high,mgsm,89.3,https://llm-stats.com/benchmarks/mgsm,"{""effort"":""default (o1 base, not high)""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=o1, slug=o1-2024-12-17, provider=OpenAI"
gpt-4-turbo,mgsm,88.5,https://llm-stats.com/benchmarks/mgsm,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-4 Turbo, slug=gpt-4-turbo-2024-04-09, provider=OpenAI"
gemini-1.5-pro,mgsm,87.5,https://llm-stats.com/benchmarks/mgsm,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemini 2.0 Flash Model Card (April 15 2025): non-thinking generation, default sampling. Same canonical setting applies to 1.5 / 2.0 family."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 1.5 Pro, slug=gemini-1.5-pro, provider=Google"
claude-3.5-haiku,mgsm,85.6,https://llm-stats.com/benchmarks/mgsm,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3.5 Haiku, slug=claude-3-5-haiku-20241022, provider=Anthropic"
qwen3-235b,mgsm,83.5,https://llm-stats.com/benchmarks/mgsm,"{""context"":""default"",""effort"":""default"",""harness"":""official (Qwen-Agent / sglang)"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Qwen3 tech report (arxiv:2505.09388) Table 11 Thinking column. Output length 32768 (38912 for AIME)."",""prompt_style"":""default (enable_thinking=True)"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3 235B A22B, slug=qwen3-235b-a22b, provider=Alibaba Cloud / Qwen Team"
claude-3-sonnet,mgsm,83.5,https://llm-stats.com/benchmarks/mgsm,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3 Sonnet, slug=claude-3-sonnet-20240229, provider=Anthropic"
gemini-1.5-flash,mgsm,82.6,https://llm-stats.com/benchmarks/mgsm,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemini 2.0 Flash Model Card (April 15 2025): non-thinking generation, default sampling. Same canonical setting applies to 1.5 / 2.0 family."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 1.5 Flash, slug=gemini-1.5-flash, provider=Google"
claude-3-haiku,mgsm,75.1,https://llm-stats.com/benchmarks/mgsm,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3 Haiku, slug=claude-3-haiku-20240307, provider=Anthropic"
phi-3.5-moe-instruct,mgsm,58.7,https://llm-stats.com/benchmarks/mgsm,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Phi-3.5-MoE-instruct, slug=phi-3.5-moe-instruct, provider=Microsoft"
gpt-3.5-turbo,mgsm,56.3,https://llm-stats.com/benchmarks/mgsm,"{""context"":""default"",""effort"":""n/a"",""harness"":""OLMES (Allen AI)"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OLMo 2 paper Table 7 Instruct (instruction-tuned variant)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-3.5 Turbo, slug=gpt-3.5-turbo-0125, provider=OpenAI"
phi-3.5-mini-instruct,mgsm,47.9,https://llm-stats.com/benchmarks/mgsm,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Phi-3.5-mini-instruct, slug=phi-3.5-mini-instruct, provider=Microsoft"
qwen3.6-plus,mmlongbench_doc,62.0,https://llm-stats.com/benchmarks/mmlongbench-doc,{},True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3.6 Plus, slug=qwen3.6-plus, provider=Alibaba Cloud / Qwen Team"
minimax-m2.1,mmlu_pro,88.0,https://llm-stats.com/benchmarks/mmlu-pro,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per MiniMax M2.5 model card (referenced as predecessor)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=MiniMax M2.1, slug=minimax-m2.1, provider=MiniMax"
longcat-flash-chat,mmlu_pro,82.7,https://llm-stats.com/benchmarks/mmlu-pro,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=LongCat-Flash-Chat, slug=longcat-flash-chat, provider=Meituan"
sarvam-30b,mmlu_pro,80.0,https://llm-stats.com/benchmarks/mmlu-pro,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Sarvam-30B, slug=sarvam-30b, provider=Sarvam AI"
grok-2,mmlu_pro,75.5,https://llm-stats.com/benchmarks/mmlu-pro,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Grok-2, slug=grok-2, provider=xAI"
grok-2-mini,mmlu_pro,72.0,https://llm-stats.com/benchmarks/mmlu-pro,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Grok-2 mini, slug=grok-2-mini, provider=xAI"
qwen-2.5-32b,mmlu_pro,69.0,https://llm-stats.com/benchmarks/mmlu-pro,"{""context"":""default"",""effort"":""n/a"",""harness"":""OLMES (Allen AI)"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OLMo 2 paper Table 7 Instruct (instruction-tuned variant)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5 32B Instruct, slug=qwen-2.5-32b-instruct, provider=Alibaba Cloud / Qwen Team"
claude-3-opus,mmlu_pro,68.5,https://llm-stats.com/benchmarks/mmlu-pro,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3 Opus, slug=claude-3-opus-20240229, provider=Anthropic"
minicpm-sala,mmlu_pro,67.0,https://llm-stats.com/benchmarks/mmlu-pro,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=MiniCPM-SALA, slug=minicpm-sala, provider=OpenBMB"
llama-3.1-70b,mmlu_pro,66.4,https://llm-stats.com/benchmarks/mmlu-pro,"{""context"":""default"",""effort"":""n/a"",""harness"":""OLMES (Allen AI)"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OLMo 2 paper Table 7 Instruct (instruction-tuned variant)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Llama 3.1 70B Instruct, slug=llama-3.1-70b-instruct, provider=Meta"
mistral-small-24b,mmlu_pro,66.3,https://llm-stats.com/benchmarks/mmlu-pro,"{""context"":""default"",""effort"":""n/a"",""harness"":""OLMES (Allen AI)"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OLMo 2 paper Table 7 Instruct (instruction-tuned variant)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Mistral Small 3 24B Instruct, slug=mistral-small-24b-instruct-2501, provider=Mistral AI"
claude-3.5-haiku,mmlu_pro,65.0,https://llm-stats.com/benchmarks/mmlu-pro,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3.5 Haiku, slug=claude-3-5-haiku-20241022, provider=Anthropic"
qwen2-72b-instruct,mmlu_pro,64.4,https://llm-stats.com/benchmarks/mmlu-pro,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2 72B Instruct, slug=qwen2-72b-instruct, provider=Alibaba Cloud / Qwen Team"
claude-3-sonnet,mmlu_pro,56.8,https://llm-stats.com/benchmarks/mmlu-pro,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3 Sonnet, slug=claude-3-sonnet-20240229, provider=Anthropic"
qwen-2.5-7b,mmlu_pro,56.3,https://llm-stats.com/benchmarks/mmlu-pro,"{""context"":""default"",""effort"":""n/a"",""harness"":""OLMES (Allen AI)"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OLMo 2 paper Table 7 Instruct (instruction-tuned variant)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5 7B Instruct, slug=qwen-2.5-7b-instruct, provider=Alibaba Cloud / Qwen Team"
grok-1.5,mmlu_pro,51.0,https://llm-stats.com/benchmarks/mmlu-pro,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Grok-1.5, slug=grok-1.5, provider=xAI"
phi-3.5-mini-instruct,mmlu_pro,47.4,https://llm-stats.com/benchmarks/mmlu-pro,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Phi-3.5-mini-instruct, slug=phi-3.5-mini-instruct, provider=Microsoft"
qwen2.5-omni-7b,mmlu_pro,47.0,https://llm-stats.com/benchmarks/mmlu-pro,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5-Omni-7B, slug=qwen2.5-omni-7b, provider=Alibaba Cloud / Qwen Team"
phi-3.5-moe-instruct,mmlu_pro,45.3,https://llm-stats.com/benchmarks/mmlu-pro,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Phi-3.5-MoE-instruct, slug=phi-3.5-moe-instruct, provider=Microsoft"
qwen2-7b-instruct,mmlu_pro,44.1,https://llm-stats.com/benchmarks/mmlu-pro,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2 7B Instruct, slug=qwen2-7b-instruct, provider=Alibaba Cloud / Qwen Team"
qwen3.6-plus,mmlu_redux,94.5,https://llm-stats.com/benchmarks/mmlu-redux,{},True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3.6 Plus, slug=qwen3.6-plus, provider=Alibaba Cloud / Qwen Team"
qwen-2.5-32b,mmlu_redux,83.9,https://llm-stats.com/benchmarks/mmlu-redux,"{""context"":""default"",""effort"":""n/a"",""harness"":""OLMES (Allen AI)"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OLMo 2 paper Table 7 Instruct (instruction-tuned variant)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5 32B Instruct, slug=qwen-2.5-32b-instruct, provider=Alibaba Cloud / Qwen Team"
mistral-large-3,mmlu_redux,82.0,https://llm-stats.com/benchmarks/mmlu-redux,{},True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Mistral Large 3, slug=mistral-large-3-2509, provider=Mistral AI"
qwen-2.5-14b-instruct,mmlu_redux,80.0,https://llm-stats.com/benchmarks/mmlu-redux,"{""context"":""default (128K)"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Phi-4 paper Table 1 baseline."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5 14B Instruct, slug=qwen-2.5-14b-instruct, provider=Alibaba Cloud / Qwen Team"
qwen-2.5-7b,mmlu_redux,75.4,https://llm-stats.com/benchmarks/mmlu-redux,"{""context"":""default"",""effort"":""n/a"",""harness"":""OLMES (Allen AI)"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OLMo 2 paper Table 7 Instruct (instruction-tuned variant)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5 7B Instruct, slug=qwen-2.5-7b-instruct, provider=Alibaba Cloud / Qwen Team"
qwen2.5-omni-7b,mmlu_redux,71.0,https://llm-stats.com/benchmarks/mmlu-redux,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5-Omni-7B, slug=qwen2.5-omni-7b, provider=Alibaba Cloud / Qwen Team"
longcat-flash-chat,mmlu,89.7,https://llm-stats.com/benchmarks/mmlu,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=LongCat-Flash-Chat, slug=longcat-flash-chat, provider=Meituan"
grok-2,mmlu,87.5,https://llm-stats.com/benchmarks/mmlu,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Grok-2, slug=grok-2, provider=xAI"
claude-3-opus,mmlu,86.8,https://llm-stats.com/benchmarks/mmlu,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3 Opus, slug=claude-3-opus-20240229, provider=Anthropic"
gpt-4-turbo,mmlu,86.5,https://llm-stats.com/benchmarks/mmlu,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-4 Turbo, slug=gpt-4-turbo-2024-04-09, provider=OpenAI"
grok-2-mini,mmlu,86.2,https://llm-stats.com/benchmarks/mmlu,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Grok-2 mini, slug=grok-2-mini, provider=xAI"
gemini-1.5-pro,mmlu,85.9,https://llm-stats.com/benchmarks/mmlu,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemini 2.0 Flash Model Card (April 15 2025): non-thinking generation, default sampling. Same canonical setting applies to 1.5 / 2.0 family."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 1.5 Pro, slug=gemini-1.5-pro, provider=Google"
o1-mini,mmlu,85.2,https://llm-stats.com/benchmarks/mmlu,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Phi-4-reasoning paper Table 1 baseline (Microsoft self-test)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=o1-mini, slug=o1-mini, provider=OpenAI"
sarvam-30b,mmlu,85.1,https://llm-stats.com/benchmarks/mmlu,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Sarvam-30B, slug=sarvam-30b, provider=Sarvam AI"
qwen2-72b-instruct,mmlu,82.3,https://llm-stats.com/benchmarks/mmlu,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2 72B Instruct, slug=qwen2-72b-instruct, provider=Alibaba Cloud / Qwen Team"
grok-1.5,mmlu,81.3,https://llm-stats.com/benchmarks/mmlu,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Grok-1.5, slug=grok-1.5, provider=xAI"
nova-lite,mmlu,80.5,https://llm-stats.com/benchmarks/mmlu,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Nova Lite, slug=nova-lite, provider=Amazon"
claude-3-sonnet,mmlu,79.0,https://llm-stats.com/benchmarks/mmlu,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3 Sonnet, slug=claude-3-sonnet-20240229, provider=Anthropic"
gemini-1.5-flash,mmlu,78.9,https://llm-stats.com/benchmarks/mmlu,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemini 2.0 Flash Model Card (April 15 2025): non-thinking generation, default sampling. Same canonical setting applies to 1.5 / 2.0 family."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 1.5 Flash, slug=gemini-1.5-flash, provider=Google"
phi-3.5-moe-instruct,mmlu,78.9,https://llm-stats.com/benchmarks/mmlu,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Phi-3.5-MoE-instruct, slug=phi-3.5-moe-instruct, provider=Microsoft"
nova-micro,mmlu,77.6,https://llm-stats.com/benchmarks/mmlu,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Nova Micro, slug=nova-micro, provider=Amazon"
claude-3-haiku,mmlu,75.2,https://llm-stats.com/benchmarks/mmlu,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3 Haiku, slug=claude-3-haiku-20240307, provider=Anthropic"
qwen2-7b-instruct,mmlu,70.5,https://llm-stats.com/benchmarks/mmlu,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2 7B Instruct, slug=qwen2-7b-instruct, provider=Alibaba Cloud / Qwen Team"
pixtral-12b,mmlu,69.2,https://llm-stats.com/benchmarks/mmlu,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Pixtral-12B, slug=pixtral-12b-2409, provider=Mistral AI"
phi-3.5-mini-instruct,mmlu,69.0,https://llm-stats.com/benchmarks/mmlu,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Phi-3.5-mini-instruct, slug=phi-3.5-mini-instruct, provider=Microsoft"
granite-3.3-8b-instruct,mmlu,65.5,https://llm-stats.com/benchmarks/mmlu,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Granite 3.3 8B Instruct, slug=granite-3.3-8b-instruct, provider=IBM"
granite-3.3-8b-base,mmlu,63.9,https://llm-stats.com/benchmarks/mmlu,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Granite 3.3 8B Base, slug=granite-3.3-8b-base, provider=IBM"
qwen3.6-plus,mmmlu,89.5,https://llm-stats.com/benchmarks/mmmlu,{},True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3.6 Plus, slug=qwen3.6-plus, provider=Alibaba Cloud / Qwen Team"
mistral-large-3,mmmlu,74.2,https://llm-stats.com/benchmarks/mmmlu,{},True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Mistral Large 3, slug=mistral-large-3-2509, provider=Mistral AI"
phi-3.5-moe-instruct,mmmlu,69.9,https://llm-stats.com/benchmarks/mmmlu,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Phi-3.5-MoE-instruct, slug=phi-3.5-moe-instruct, provider=Microsoft"
phi-3.5-mini-instruct,mmmlu,55.4,https://llm-stats.com/benchmarks/mmmlu,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Phi-3.5-mini-instruct, slug=phi-3.5-mini-instruct, provider=Microsoft"
qwen3.6-plus,mmmu_pro,78.8,https://llm-stats.com/benchmarks/mmmu-pro,{},True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3.6 Plus, slug=qwen3.6-plus, provider=Alibaba Cloud / Qwen Team"
gpt-4o,mmmu_pro,59.9,https://llm-stats.com/benchmarks/mmmu-pro,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OpenAI GPT-4.1 blog table."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-4o, slug=gpt-4o-2024-08-06, provider=OpenAI"
llama-4-maverick,mmmu_pro,59.6,https://llm-stats.com/benchmarks/mmmu-pro,"{""context"":""default"",""effort"":""n/a"",""harness"":""official-meta"",""judge"":""rule-based"",""mode"":""non-thinking"",""multimodal_input"":true,""prompt_style"":""zeroshot"",""sampling"":""pass@1"",""temperature"":""0.0"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Llama 4 Maverick, slug=llama-4-maverick, provider=Meta"
qwen2.5-omni-7b,mmmu_pro,36.6,https://llm-stats.com/benchmarks/mmmu-pro,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5-Omni-7B, slug=qwen2.5-omni-7b, provider=Alibaba Cloud / Qwen Team"
qwen3.6-plus,mmmu,86.0,https://llm-stats.com/benchmarks/mmmu,{},True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3.6 Plus, slug=qwen3.6-plus, provider=Alibaba Cloud / Qwen Team"
grok-2,mmmu,66.1,https://llm-stats.com/benchmarks/mmmu,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Grok-2, slug=grok-2, provider=xAI"
grok-2-mini,mmmu,63.2,https://llm-stats.com/benchmarks/mmmu,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Grok-2 mini, slug=grok-2-mini, provider=xAI"
qwen2.5-omni-7b,mmmu,59.2,https://llm-stats.com/benchmarks/mmmu,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5-Omni-7B, slug=qwen2.5-omni-7b, provider=Alibaba Cloud / Qwen Team"
nova-lite,mmmu,56.2,https://llm-stats.com/benchmarks/mmmu,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Nova Lite, slug=nova-lite, provider=Amazon"
gemini-1.5-flash-8b,mmmu,53.7,https://llm-stats.com/benchmarks/mmmu,"{""context"":""1M"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Command A paper baseline."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 1.5 Flash 8B, slug=gemini-1.5-flash-8b, provider=Google"
grok-1.5,mmmu,53.6,https://llm-stats.com/benchmarks/mmmu,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Grok-1.5, slug=grok-1.5, provider=xAI"
pixtral-12b,mmmu,52.5,https://llm-stats.com/benchmarks/mmmu,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Pixtral-12B, slug=pixtral-12b-2409, provider=Mistral AI"
gpt-3.5-turbo,mmmu,0.0,https://llm-stats.com/benchmarks/mmmu,"{""context"":""default"",""effort"":""n/a"",""harness"":""OLMES (Allen AI)"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OLMo 2 paper Table 7 Instruct (instruction-tuned variant)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-3.5 Turbo, slug=gpt-3.5-turbo-0125, provider=OpenAI"
qwen3.6-plus,mmstar,83.3,https://llm-stats.com/benchmarks/mmstar,{},True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3.6 Plus, slug=qwen3.6-plus, provider=Alibaba Cloud / Qwen Team"
qwen2.5-omni-7b,mmstar,64.0,https://llm-stats.com/benchmarks/mmstar,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5-Omni-7B, slug=qwen2.5-omni-7b, provider=Alibaba Cloud / Qwen Team"
deepseek-v4-pro,mrcr_v2_2needle_1m,83.5,https://llm-stats.com/benchmarks/mrcr-1m,"{""effort"":""max""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek-V4-Pro-Max, slug=deepseek-v4-pro-max, provider=DeepSeek"
deepseek-v4-flash,mrcr_v2_2needle_1m,78.7,https://llm-stats.com/benchmarks/mrcr-1m,"{""effort"":""max""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek-V4-Flash-Max, slug=deepseek-v4-flash-max, provider=DeepSeek"
gemini-2.0-flash-lite,mrcr_v2_2needle_1m,58.0,https://llm-stats.com/benchmarks/mrcr-1m,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemini 2.0 Flash Model Card (April 15 2025): non-thinking generation, default sampling. Same canonical setting applies to 1.5 / 2.0 family."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 2.0 Flash-Lite, slug=gemini-2.0-flash-lite, provider=Google"
claude-opus-4.6,mrcr_v2_8needle,93.0,https://llm-stats.com/benchmarks/mrcr-v2-(8-needle),"{""context"":""up to 1M"",""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Mythos System Card Table 6.3.A: standard config = adaptive thinking max effort, default sampling, avg 5 trials, context up to 1M."",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude Opus 4.6, slug=claude-opus-4-6, provider=Anthropic"
gpt-5.5,mrcr_v2_8needle,74.0,https://llm-stats.com/benchmarks/mrcr-v2-(8-needle),"{""context"":""default"",""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per OpenAI gpt-5.5 blog: reasoning effort=xhigh, research environment (may differ from production ChatGPT)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-5.5, slug=gpt-5.5, provider=OpenAI"
gemini-3.1-pro,mrcr_v2_8needle,26.3,https://llm-stats.com/benchmarks/mrcr-v2-(8-needle),"{""context"":""default"",""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Google DeepMind /models/gemini/pro/ table: Gemini 3.1 Pro Thinking (High), no tools by default."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 3.1 Pro, slug=gemini-3.1-pro-preview, provider=Google"
gemini-3-flash,mrcr_v2_8needle,22.1,https://llm-stats.com/benchmarks/mrcr-v2-(8-needle),"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Google DeepMind /models/gemini/flash/ table: Gemini 3 Flash Thinking, no tools by default."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 3 Flash, slug=gemini-3-flash-preview, provider=Google"
gemini-2.5-pro,mrcr_v2_8needle,16.4,https://llm-stats.com/benchmarks/mrcr-v2-(8-needle),"{""context"":""default"",""effort"":""default"",""harness"":""AI Studio API"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Gemini 2.5 Pro Model Card (GA, June 27 2025): pass@1 single attempt, AI Studio API model-id gemini-2.5-pro-preview-06-05 / gemini-2.5-pro GA, default sampling. Release date set to GA date 2025-06-27."",""prompt_style"":""default"",""sampling"":""pass@1 (single attempt)"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 2.5 Pro Preview 06-05, slug=gemini-2.5-pro-preview-06-05, provider=Google"
gemini-2.5-pro,mrcr_v1,93.0,https://llm-stats.com/benchmarks/mrcr,"{""context"":""default"",""effort"":""default"",""harness"":""AI Studio API"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Gemini 2.5 Pro Model Card (GA, June 27 2025): pass@1 single attempt, AI Studio API model-id gemini-2.5-pro-preview-06-05 / gemini-2.5-pro GA, default sampling. Release date set to GA date 2025-06-27."",""prompt_style"":""default"",""sampling"":""pass@1 (single attempt)"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 2.5 Pro, slug=gemini-2.5-pro, provider=Google"
gemini-1.5-flash-8b,mrcr_v1,54.7,https://llm-stats.com/benchmarks/mrcr,"{""context"":""1M"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Command A paper baseline."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 1.5 Flash 8B, slug=gemini-1.5-flash-8b, provider=Google"
deepseek-v2.5-0905,mt_bench_101,90.2,https://llm-stats.com/benchmarks/mt-bench,"{""snapshot"":""0905""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek-V2.5, slug=deepseek-v2.5, provider=DeepSeek"
qwen-2.5-7b,mt_bench_101,87.5,https://llm-stats.com/benchmarks/mt-bench,"{""context"":""default"",""effort"":""n/a"",""harness"":""OLMES (Allen AI)"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OLMo 2 paper Table 7 Instruct (instruction-tuned variant)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5 7B Instruct, slug=qwen-2.5-7b-instruct, provider=Alibaba Cloud / Qwen Team"
mistral-large-2,mt_bench_101,86.3,https://llm-stats.com/benchmarks/mt-bench,"{""context"":""128K"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Command A paper (third-party Cohere reproduction baseline)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Mistral Large 2, slug=mistral-large-2-2407, provider=Mistral AI"
qwen2-7b-instruct,mt_bench_101,84.1,https://llm-stats.com/benchmarks/mt-bench,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2 7B Instruct, slug=qwen2-7b-instruct, provider=Alibaba Cloud / Qwen Team"
mistral-small-24b,mt_bench_101,83.5,https://llm-stats.com/benchmarks/mt-bench,"{""context"":""default"",""effort"":""n/a"",""harness"":""OLMES (Allen AI)"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OLMo 2 paper Table 7 Instruct (instruction-tuned variant)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Mistral Small 3 24B Instruct, slug=mistral-small-24b-instruct-2501, provider=Mistral AI"
pixtral-12b,mt_bench_101,76.8,https://llm-stats.com/benchmarks/mt-bench,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Pixtral-12B, slug=pixtral-12b-2409, provider=Mistral AI"
qwen2.5-omni-7b,muirbench,59.2,https://llm-stats.com/benchmarks/muirbench,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5-Omni-7B, slug=qwen2.5-omni-7b, provider=Alibaba Cloud / Qwen Team"
minimax-m2.5,multi_swe_bench,51.3,https://llm-stats.com/benchmarks/multi-swe-bench,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per MiniMax M2.5 model card."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=MiniMax M2.5, slug=minimax-m2.5, provider=MiniMax"
minimax-m2.1,multi_swe_bench,49.4,https://llm-stats.com/benchmarks/multi-swe-bench,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per MiniMax M2.5 model card (referenced as predecessor)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=MiniMax M2.1, slug=minimax-m2.1, provider=MiniMax"
gpt-5,multichallenge,69.6,https://llm-stats.com/benchmarks/multichallenge,"{""context"":""default"",""effort"":""high"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per OpenAI GPT-5 dev blog. high reasoning effort."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-5, slug=gpt-5-2025-08-07, provider=OpenAI"
o3-high,multichallenge,60.4,https://llm-stats.com/benchmarks/multichallenge,"{""effort"":""default (o3 base, not high)""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=o3, slug=o3-2025-04-16, provider=OpenAI"
kimi-k2,multichallenge,54.1,https://llm-stats.com/benchmarks/multichallenge,"{""snapshot"":""0905"",""variant"":""instruct""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Kimi K2-Instruct-0905, slug=kimi-k2-instruct-0905, provider=Moonshot AI"
o4-mini-high,multichallenge,43.0,https://llm-stats.com/benchmarks/multichallenge,"{""effort"":""default (o4-mini base, not high)""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=o4-mini, slug=o4-mini, provider=OpenAI"
qwen3-235b,multipl_e_avg,87.9,https://llm-stats.com/benchmarks/multipl-e,"{""mode"":""instruct"",""snapshot"":""2507""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3-235B-A22B-Instruct-2507, slug=qwen3-235b-a22b-instruct-2507, provider=Alibaba Cloud / Qwen Team"
kimi-k2,multipl_e_avg,85.7,https://llm-stats.com/benchmarks/multipl-e,"{""variant"":""instruct (BP kimi-k2 is base)""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Kimi K2 Instruct, slug=kimi-k2-instruct, provider=Moonshot AI"
qwen-2.5-32b,multipl_e_avg,75.4,https://llm-stats.com/benchmarks/multipl-e,"{""context"":""default"",""effort"":""n/a"",""harness"":""OLMES (Allen AI)"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OLMo 2 paper Table 7 Instruct (instruction-tuned variant)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5 32B Instruct, slug=qwen-2.5-32b-instruct, provider=Alibaba Cloud / Qwen Team"
qwen-2.5-14b-instruct,multipl_e_avg,72.8,https://llm-stats.com/benchmarks/multipl-e,"{""context"":""default (128K)"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Phi-4 paper Table 1 baseline."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5 14B Instruct, slug=qwen-2.5-14b-instruct, provider=Alibaba Cloud / Qwen Team"
qwen-2.5-7b,multipl_e_avg,70.4,https://llm-stats.com/benchmarks/multipl-e,"{""context"":""default"",""effort"":""n/a"",""harness"":""OLMES (Allen AI)"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OLMo 2 paper Table 7 Instruct (instruction-tuned variant)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5 7B Instruct, slug=qwen-2.5-7b-instruct, provider=Alibaba Cloud / Qwen Team"
qwen2-72b-instruct,multipl_e_avg,69.2,https://llm-stats.com/benchmarks/multipl-e,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2 72B Instruct, slug=qwen2-72b-instruct, provider=Alibaba Cloud / Qwen Team"
qwen2.5-omni-7b,multipl_e_avg,65.8,https://llm-stats.com/benchmarks/multipl-e,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5-Omni-7B, slug=qwen2.5-omni-7b, provider=Alibaba Cloud / Qwen Team"
qwen2-7b-instruct,multipl_e_avg,59.1,https://llm-stats.com/benchmarks/multipl-e,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2 7B Instruct, slug=qwen2-7b-instruct, provider=Alibaba Cloud / Qwen Team"
qwen2.5-omni-7b,ocrbench_v2,57.8,https://llm-stats.com/benchmarks/ocrbench-v2,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5-Omni-7B, slug=qwen2.5-omni-7b, provider=Alibaba Cloud / Qwen Team"
qwen3-235b,ojbench,32.5,https://llm-stats.com/benchmarks/ojbench,"{""mode"":""thinking"",""snapshot"":""2507""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3-235B-A22B-Thinking-2507, slug=qwen3-235b-a22b-thinking-2507, provider=Alibaba Cloud / Qwen Team"
qwen3.6-plus,omnidocbench_1.5,91.2,https://llm-stats.com/benchmarks/omnidocbench-1.5,{},True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3.6 Plus, slug=qwen3.6-plus, provider=Alibaba Cloud / Qwen Team"
gpt-5.4,omnidocbench_1.5,89.1,https://llm-stats.com/benchmarks/omnidocbench-1.5,"{""context"":""default"",""effort"":""xhigh"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per OpenAI gpt-5.4 blog: reasoning effort=xhigh, research environment (may differ from production ChatGPT)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-5.4, slug=gpt-5.4, provider=OpenAI"
granite-3.3-8b-base,popqa,26.2,https://llm-stats.com/benchmarks/popqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Granite 3.3 8B Base, slug=granite-3.3-8b-base, provider=IBM"
granite-3.3-8b-instruct,popqa,26.2,https://llm-stats.com/benchmarks/popqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Granite 3.3 8B Instruct, slug=granite-3.3-8b-instruct, provider=IBM"
qwen3.6-plus,realworldqa,85.4,https://llm-stats.com/benchmarks/realworldqa,{},True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3.6 Plus, slug=qwen3.6-plus, provider=Alibaba Cloud / Qwen Team"
qwen2.5-omni-7b,realworldqa,70.3,https://llm-stats.com/benchmarks/realworldqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5-Omni-7B, slug=qwen2.5-omni-7b, provider=Alibaba Cloud / Qwen Team"
phi-3.5-moe-instruct,repoqa,85.0,https://llm-stats.com/benchmarks/repoqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Phi-3.5-MoE-instruct, slug=phi-3.5-moe-instruct, provider=Microsoft"
phi-3.5-mini-instruct,repoqa,77.0,https://llm-stats.com/benchmarks/repoqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Phi-3.5-mini-instruct, slug=phi-3.5-mini-instruct, provider=Microsoft"
minicpm-sala,ruler_128k,89.4,https://llm-stats.com/benchmarks/ruler-128k,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=MiniCPM-SALA, slug=minicpm-sala, provider=OpenBMB"
qwen3.6-plus,screenspot_pro,68.2,https://llm-stats.com/benchmarks/screenspot-pro,{},True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3.6 Plus, slug=qwen3.6-plus, provider=Alibaba Cloud / Qwen Team"
kimi-k2-thinking,seal_0,56.3,https://llm-stats.com/benchmarks/seal-0,"{""context"":""default (256K)"",""effort"":""default (96k thinking-token budget for HLE/AIME/HMMT/GPQA; 128k for IMO/LCB/OJ)"",""harness"":""official"",""judge"":""rule-based (HLE uses o3-mini)"",""mode"":""thinking"",""notes"":""Per Kimi K2 Thinking blog (moonshotai.github.io/Kimi-K2/thinking.html). Thinking model."",""prompt_style"":""default"",""sampling"":""pass@1 (avg@32 for AIME/HMMT no-tools)"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Kimi K2-Thinking-0905, slug=kimi-k2-thinking-0905, provider=Moonshot AI"
deepseek-v4-pro,simpleqa,57.9,https://llm-stats.com/benchmarks/simpleqa,"{""effort"":""max""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek-V4-Pro-Max, slug=deepseek-v4-pro-max, provider=DeepSeek"
o1-high,simpleqa,47.0,https://llm-stats.com/benchmarks/simpleqa,"{""effort"":""default (o1 base, not high)""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=o1, slug=o1-2024-12-17, provider=OpenAI"
gpt-4o,simpleqa,38.2,https://llm-stats.com/benchmarks/simpleqa,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OpenAI GPT-4.1 blog table."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-4o, slug=gpt-4o-2024-08-06, provider=OpenAI"
deepseek-v4-flash,simpleqa,34.1,https://llm-stats.com/benchmarks/simpleqa,"{""effort"":""max""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek-V4-Flash-Max, slug=deepseek-v4-flash-max, provider=DeepSeek"
gemma-3-1b-it,simpleqa,2.2,https://llm-stats.com/benchmarks/simpleqa,"{""context"":""32k"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Google Gemma 3 1B instruction-tuned."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemma 3 1B, slug=gemma-3-1b-it, provider=Google"
qwen3.6-plus,simplevqa,67.3,https://llm-stats.com/benchmarks/simplevqa,{},True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3.6 Plus, slug=qwen3.6-plus, provider=Alibaba Cloud / Qwen Team"
qwen3.6-plus,supergpqa,71.6,https://llm-stats.com/benchmarks/supergpqa,{},True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3.6 Plus, slug=qwen3.6-plus, provider=Alibaba Cloud / Qwen Team"
qwen3-235b,supergpqa,64.9,https://llm-stats.com/benchmarks/supergpqa,"{""mode"":""thinking"",""snapshot"":""2507""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3-235B-A22B-Thinking-2507, slug=qwen3-235b-a22b-thinking-2507, provider=Alibaba Cloud / Qwen Team"
qwen3.6-plus,swe_bench_multilingual,73.8,https://llm-stats.com/benchmarks/swe-bench-multilingual,{},True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3.6 Plus, slug=qwen3.6-plus, provider=Alibaba Cloud / Qwen Team"
minimax-m2.1,swe_bench_multilingual,72.5,https://llm-stats.com/benchmarks/swe-bench-multilingual,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per MiniMax M2.5 model card (referenced as predecessor)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=MiniMax M2.1, slug=minimax-m2.1, provider=MiniMax"
deepseek-r1,swe_bench_multilingual,30.5,https://llm-stats.com/benchmarks/swe-bench-multilingual,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per DeepSeek R1-0528 model card."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek-R1-0528, slug=deepseek-r1-0528, provider=DeepSeek"
minimax-m2.5,swe_bench_pro,55.4,https://llm-stats.com/benchmarks/swe-bench-pro,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per MiniMax M2.5 model card."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=MiniMax M2.5, slug=minimax-m2.5, provider=MiniMax"
minimax-m2.5,swe_bench_verified,80.2,https://llm-stats.com/benchmarks/swe-bench-verified,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per MiniMax M2.5 model card."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=MiniMax M2.5, slug=minimax-m2.5, provider=MiniMax"
minimax-m2.1,swe_bench_verified,67.0,https://llm-stats.com/benchmarks/swe-bench-verified,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per MiniMax M2.5 model card (referenced as predecessor)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=MiniMax M2.1, slug=minimax-m2.1, provider=MiniMax"
longcat-flash-chat,swe_bench_verified,60.4,https://llm-stats.com/benchmarks/swe-bench-verified,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=LongCat-Flash-Chat, slug=longcat-flash-chat, provider=Meituan"
claude-3.5-haiku,swe_bench_verified,40.6,https://llm-stats.com/benchmarks/swe-bench-verified,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3.5 Haiku, slug=claude-3-5-haiku-20241022, provider=Anthropic"
sarvam-30b,swe_bench_verified,34.0,https://llm-stats.com/benchmarks/swe-bench-verified,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Sarvam-30B, slug=sarvam-30b, provider=Sarvam AI"
gemini-diffusion,swe_bench_verified,22.9,https://llm-stats.com/benchmarks/swe-bench-verified,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini Diffusion, slug=gemini-diffusion, provider=Google"
gpt-4.5,swelancer,37.3,https://llm-stats.com/benchmarks/swe-lancer,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OpenAI GPT-4.1 blog table."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-4.5, slug=gpt-4.5, provider=OpenAI"
gpt-4o,swelancer,32.6,https://llm-stats.com/benchmarks/swe-lancer,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OpenAI GPT-4.1 blog table."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-4o, slug=gpt-4o-2024-08-06, provider=OpenAI"
claude-sonnet-4.5,tau_bench_airline,70.0,https://llm-stats.com/benchmarks/tau-bench-airline,"{""context"":""200K (default)"",""effort"":""high (default)"",""harness"":""Terminus-2 (for terminal-bench); official otherwise"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per Anthropic Opus 4.5 announcement: 64K thinking budget, interleaved scratchpads, 200K context, default effort=high, default sampling, avg 5 trials. Exceptions: SWE-bench Verified=no thinking; Terminal-Bench=128K thinking budget."",""prompt_style"":""interleaved scratchpads"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""default (top_p, temperature)"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude Sonnet 4.5, slug=claude-sonnet-4-5-20250929, provider=Anthropic"
o4-mini-high,tau_bench_airline,49.2,https://llm-stats.com/benchmarks/tau-bench-airline,"{""effort"":""default (o4-mini base, not high)""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=o4-mini, slug=o4-mini, provider=OpenAI"
qwen3-235b,tau_bench_airline,46.0,https://llm-stats.com/benchmarks/tau-bench-airline,"{""mode"":""thinking"",""snapshot"":""2507""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3-235B-A22B-Thinking-2507, slug=qwen3-235b-a22b-thinking-2507, provider=Alibaba Cloud / Qwen Team"
claude-3.5-haiku,tau_bench_airline,22.8,https://llm-stats.com/benchmarks/tau-bench-airline,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3.5 Haiku, slug=claude-3-5-haiku-20241022, provider=Anthropic"
o4-mini-high,tau_bench_retail,71.8,https://llm-stats.com/benchmarks/tau-bench-retail,"{""effort"":""default (o4-mini base, not high)""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=o4-mini, slug=o4-mini, provider=OpenAI"
qwen3-235b,tau_bench_retail,67.8,https://llm-stats.com/benchmarks/tau-bench-retail,"{""mode"":""thinking"",""snapshot"":""2507""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3-235B-A22B-Thinking-2507, slug=qwen3-235b-a22b-thinking-2507, provider=Alibaba Cloud / Qwen Team"
claude-3.5-haiku,tau_bench_retail,51.0,https://llm-stats.com/benchmarks/tau-bench-retail,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Claude 3.5 Haiku, slug=claude-3-5-haiku-20241022, provider=Anthropic"
glm-4.7,tau1_bench_avg,87.4,https://llm-stats.com/benchmarks/tau-bench,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per GLM-5 blog."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GLM-4.7, slug=glm-4.7, provider=Zhipu AI"
minimax-m2,tau1_bench_avg,77.2,https://llm-stats.com/benchmarks/tau-bench,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per MiniMax M2 model card."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=MiniMax M2, slug=minimax-m2, provider=MiniMax"
o3-high,tau1_bench_avg,63.0,https://llm-stats.com/benchmarks/tau-bench,"{""effort"":""default (o3 base, not high)""}",False,third_party_aggregator,verified_third_party,"llm-stats.com row: model=o3, slug=o3-2025-04-16, provider=OpenAI"
deepseek-v3.2-speciale,terminal_bench,46.4,https://llm-stats.com/benchmarks/terminal-bench-2,"{""context"":""128K"",""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""DS V3.2-Speciale: high-compute reasoning variant of V3.2. Per V3.2 tech report Table 3."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek-V3.2-Speciale, slug=deepseek-v3.2-speciale, provider=DeepSeek"
minimax-m2.1,toolathlon,43.5,https://llm-stats.com/benchmarks/toolathlon,"{""context"":""default"",""effort"":""default"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per MiniMax M2.5 model card (referenced as predecessor)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=MiniMax M2.1, slug=minimax-m2.1, provider=MiniMax"
deepseek-v3.2-speciale,toolathlon,35.2,https://llm-stats.com/benchmarks/toolathlon,"{""context"":""128K"",""effort"":""max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""DS V3.2-Speciale: high-compute reasoning variant of V3.2. Per V3.2 tech report Table 3."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=DeepSeek-V3.2-Speciale, slug=deepseek-v3.2-speciale, provider=DeepSeek"
phi-3.5-moe-instruct,truthfulqa,77.5,https://llm-stats.com/benchmarks/truthfulqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Phi-3.5-MoE-instruct, slug=phi-3.5-moe-instruct, provider=Microsoft"
granite-3.3-8b-instruct,truthfulqa,66.9,https://llm-stats.com/benchmarks/truthfulqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Granite 3.3 8B Instruct, slug=granite-3.3-8b-instruct, provider=IBM"
phi-4-mini,truthfulqa,66.4,https://llm-stats.com/benchmarks/truthfulqa,"{""context"":""default (128K)"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Phi-4-mini-instruct HF model card."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Phi 4 Mini, slug=phi-4-mini, provider=Microsoft"
phi-3.5-mini-instruct,truthfulqa,64.0,https://llm-stats.com/benchmarks/truthfulqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Phi-3.5-mini-instruct, slug=phi-3.5-mini-instruct, provider=Microsoft"
qwen-2.5-14b-instruct,truthfulqa,58.4,https://llm-stats.com/benchmarks/truthfulqa,"{""context"":""default (128K)"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Phi-4 paper Table 1 baseline."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2.5 14B Instruct, slug=qwen-2.5-14b-instruct, provider=Alibaba Cloud / Qwen Team"
qwen2-72b-instruct,truthfulqa,54.8,https://llm-stats.com/benchmarks/truthfulqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen2 72B Instruct, slug=qwen2-72b-instruct, provider=Alibaba Cloud / Qwen Team"
granite-3.3-8b-base,truthfulqa,52.1,https://llm-stats.com/benchmarks/truthfulqa,"{""context"":""unknown"",""effort"":""n/a"",""harness"":""unknown"",""judge"":""unknown"",""mode"":""unknown"",""notes"":""Model added from third-party aggregator (llm-stats.com); canonical setting not established."",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unknown""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Granite 3.3 8B Base, slug=granite-3.3-8b-base, provider=IBM"
gemini-2.0-flash,vibe_eval,56.3,https://llm-stats.com/benchmarks/vibe-eval,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemini 2.0 Flash Model Card (April 15 2025): non-thinking generation, default sampling. Same canonical setting applies to 1.5 / 2.0 family."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 2.0 Flash, slug=gemini-2.0-flash, provider=Google"
gemini-1.5-pro,vibe_eval,53.9,https://llm-stats.com/benchmarks/vibe-eval,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemini 2.0 Flash Model Card (April 15 2025): non-thinking generation, default sampling. Same canonical setting applies to 1.5 / 2.0 family."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 1.5 Pro, slug=gemini-1.5-pro, provider=Google"
gemini-1.5-flash,vibe_eval,48.9,https://llm-stats.com/benchmarks/vibe-eval,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemini 2.0 Flash Model Card (April 15 2025): non-thinking generation, default sampling. Same canonical setting applies to 1.5 / 2.0 family."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 1.5 Flash, slug=gemini-1.5-flash, provider=Google"
gemini-1.5-flash-8b,vibe_eval,40.9,https://llm-stats.com/benchmarks/vibe-eval,"{""context"":""1M"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Command A paper baseline."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 1.5 Flash 8B, slug=gemini-1.5-flash-8b, provider=Google"
qwen3.6-plus,video_mme,84.2,https://llm-stats.com/benchmarks/video-mme,{},True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3.6 Plus, slug=qwen3.6-plus, provider=Alibaba Cloud / Qwen Team"
gemini-1.5-pro,video_mme,78.6,https://llm-stats.com/benchmarks/video-mme,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemini 2.0 Flash Model Card (April 15 2025): non-thinking generation, default sampling. Same canonical setting applies to 1.5 / 2.0 family."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 1.5 Pro, slug=gemini-1.5-pro, provider=Google"
gemini-1.5-flash,video_mme,76.1,https://llm-stats.com/benchmarks/video-mme,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Gemini 2.0 Flash Model Card (April 15 2025): non-thinking generation, default sampling. Same canonical setting applies to 1.5 / 2.0 family."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 1.5 Flash, slug=gemini-1.5-flash, provider=Google"
gemini-1.5-flash-8b,video_mme,66.2,https://llm-stats.com/benchmarks/video-mme,"{""context"":""1M"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per Command A paper baseline."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Gemini 1.5 Flash 8B, slug=gemini-1.5-flash-8b, provider=Google"
qwen3.6-plus,video_mmmu,84.0,https://llm-stats.com/benchmarks/videommmu,{},True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3.6 Plus, slug=qwen3.6-plus, provider=Alibaba Cloud / Qwen Team"
gpt-4o,video_mmmu,61.2,https://llm-stats.com/benchmarks/videommmu,"{""context"":""default"",""effort"":""n/a"",""harness"":""official"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OpenAI GPT-4.1 blog table."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=GPT-4o, slug=gpt-4o-2024-08-06, provider=OpenAI"
qwen3.6-plus,vitabench,44.3,https://llm-stats.com/benchmarks/vita-bench,{},True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3.6 Plus, slug=qwen3.6-plus, provider=Alibaba Cloud / Qwen Team"
qwen3.6-plus,widesearch,74.3,https://llm-stats.com/benchmarks/widesearch,{},True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Qwen3.6 Plus, slug=qwen3.6-plus, provider=Alibaba Cloud / Qwen Team"
mistral-large-3,wildbench,68.5,https://llm-stats.com/benchmarks/wild-bench,{},True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Mistral Large 3, slug=mistral-large-3-2509, provider=Mistral AI"
mistral-small-24b,wildbench,52.2,https://llm-stats.com/benchmarks/wild-bench,"{""context"":""default"",""effort"":""n/a"",""harness"":""OLMES (Allen AI)"",""judge"":""rule-based"",""mode"":""non-thinking"",""notes"":""Per OLMo 2 paper Table 7 Instruct (instruction-tuned variant)."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,third_party_aggregator,verified_third_party,"llm-stats.com row: model=Mistral Small 3 24B Instruct, slug=mistral-small-24b-instruct-2501, provider=Mistral AI"
kimi-k2.6,kimi_code_bench_v2,50.9,https://huggingface.co/moonshotai/Kimi-K2.7-Code,"{""context"":""262144 (256K)"",""effort"":""default"",""harness"":""Kimi Code CLI + official benchmark harness"",""judge"":""Moonshot in-house evaluator"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95"",""tools"":""benchmark-specified""}",False,model_card,verified,Kimi K2.7 Code model card Evaluation Results table: 50.9. This source uses top_p=0.95; the existing Kimi K2.6 canonical setting uses top_p=1.0.
kimi-k2.7-code,kimi_code_bench_v2,62.0,https://huggingface.co/moonshotai/Kimi-K2.7-Code,"{""context"":""262144 (256K)"",""effort"":""default"",""harness"":""Kimi Code CLI + official benchmark harness"",""judge"":""Moonshot in-house evaluator"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95"",""tools"":""benchmark-specified""}",True,model_card,verified,Kimi K2.7 Code model card Evaluation Results table: 62.0.
gpt-5.5,kimi_code_bench_v2,69.0,https://huggingface.co/moonshotai/Kimi-K2.7-Code,"{""context"":""default"",""effort"":""xhigh"",""harness"":""Codex + official benchmark harness"",""judge"":""Moonshot in-house evaluator"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,model_card,verified,Kimi K2.7 Code model card Evaluation Results table: 69.0. Cross-model value reported by Moonshot AI.
claude-opus-4.8,kimi_code_bench_v2,71.7,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic coding harness per row"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: Kimi Code Bench 2.0 = 71.7. Source setting: agentic coding harness per row. Origin: Moonshot evaluation.
kimi-k2.6,program_bench,48.3,https://huggingface.co/moonshotai/Kimi-K2.7-Code,"{""context"":""262144 (256K)"",""effort"":""default"",""harness"":""Kimi Code CLI + official benchmark harness"",""judge"":""248,000+ fuzz-generated behavioral tests"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95"",""tools"":""benchmark-specified""}",False,model_card,verified,Kimi K2.7 Code model card Evaluation Results table: 48.3. This source uses top_p=0.95; the existing Kimi K2.6 canonical setting uses top_p=1.0.
kimi-k2.7-code,program_bench,53.6,https://huggingface.co/moonshotai/Kimi-K2.7-Code,"{""context"":""262144 (256K)"",""effort"":""default"",""harness"":""Kimi Code CLI + official benchmark harness"",""judge"":""248,000+ fuzz-generated behavioral tests"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95"",""tools"":""benchmark-specified""}",True,model_card,verified,Kimi K2.7 Code model card Evaluation Results table: 53.6.
gpt-5.5,program_bench,70.8,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""agentic coding harness per row"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: ProgramBench = 70.8. Source setting: agentic coding harness per row. Origin: Moonshot K3 run; competitor values from Vals AI.
claude-opus-4.8,program_bench,71.9,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic coding harness per row"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: ProgramBench = 71.9. Source setting: agentic coding harness per row. Origin: Moonshot K3 run; competitor values from Vals AI.
kimi-k2.6,mls_bench_lite,26.7,https://huggingface.co/moonshotai/Kimi-K2.7-Code,"{""context"":""262144 (256K)"",""effort"":""default"",""harness"":""Kimi Code CLI + official benchmark harness"",""judge"":""official MLS-Bench scoring"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95"",""tools"":""benchmark-specified""}",False,model_card,verified,Kimi K2.7 Code model card Evaluation Results table: 26.7. This source uses top_p=0.95; the existing Kimi K2.6 canonical setting uses top_p=1.0.
kimi-k2.7-code,mls_bench_lite,35.1,https://huggingface.co/moonshotai/Kimi-K2.7-Code,"{""context"":""262144 (256K)"",""effort"":""default"",""harness"":""Kimi Code CLI + official benchmark harness"",""judge"":""official MLS-Bench scoring"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95"",""tools"":""benchmark-specified""}",True,model_card,verified,Kimi K2.7 Code model card Evaluation Results table: 35.1.
gpt-5.5,mls_bench_lite,35.5,https://huggingface.co/moonshotai/Kimi-K2.7-Code,"{""context"":""default"",""effort"":""xhigh"",""harness"":""Codex + official benchmark harness"",""judge"":""official MLS-Bench scoring"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,model_card,verified,Kimi K2.7 Code model card Evaluation Results table: 35.5. Cross-model value reported by Moonshot AI.
claude-opus-4.8,mls_bench_lite,42.8,https://huggingface.co/moonshotai/Kimi-K2.7-Code,"{""context"":""default"",""effort"":""max"",""harness"":""Claude Code + official benchmark harness"",""judge"":""official MLS-Bench scoring"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,model_card,verified,"Kimi K2.7 Code model card Evaluation Results table: 42.8. Cross-model value reported by Moonshot AI. This source uses xhigh effort, or max for MLS-Bench-Lite; Anthropic's model default is high effort."
kimi-k2.6,kimi_claw_24_7,42.9,https://huggingface.co/moonshotai/Kimi-K2.7-Code,"{""context"":""262144 (256K)"",""effort"":""default"",""harness"":""OpenClaw"",""judge"":""evaluation-point pass rate"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""1.0; top_p=0.95"",""tools"":""benchmark-specified""}",False,model_card,verified,Kimi K2.7 Code model card Evaluation Results table: 42.9. This source uses top_p=0.95; the existing Kimi K2.6 canonical setting uses top_p=1.0.
kimi-k2.7-code,kimi_claw_24_7,46.9,https://huggingface.co/moonshotai/Kimi-K2.7-Code,"{""context"":""262144 (256K)"",""effort"":""default"",""harness"":""OpenClaw"",""judge"":""evaluation-point pass rate"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""1.0; top_p=0.95"",""tools"":""benchmark-specified""}",True,model_card,verified,Kimi K2.7 Code model card Evaluation Results table: 46.9.
gpt-5.5,kimi_claw_24_7,52.8,https://huggingface.co/moonshotai/Kimi-K2.7-Code,"{""context"":""default"",""effort"":""xhigh"",""harness"":""OpenClaw"",""judge"":""evaluation-point pass rate"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,model_card,verified,"Kimi K2.7 Code model card Evaluation Results table: 52.8. Cross-model value reported by Moonshot AI. The source averages three OpenClaw runs, while the GPT-5.5 canonical sampling is pass@1."
claude-opus-4.8,kimi_claw_24_7,50.4,https://huggingface.co/moonshotai/Kimi-K2.7-Code,"{""context"":""default"",""effort"":""xhigh"",""harness"":""OpenClaw"",""judge"":""evaluation-point pass rate"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,model_card,verified,"Kimi K2.7 Code model card Evaluation Results table: 50.4. Cross-model value reported by Moonshot AI. This source uses xhigh effort, or max for MLS-Bench-Lite; Anthropic's model default is high effort."
kimi-k2.6,mcpmark_verified,72.8,https://huggingface.co/moonshotai/Kimi-K2.7-Code,"{""context"":""262144 (256K)"",""effort"":""default"",""harness"":""official MCPMark; 100 steps; 32K max tokens per step"",""judge"":""programmatic verification on human-verified tasks"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""1.0; top_p=0.95"",""tools"":""benchmark-specified""}",False,model_card,verified,Kimi K2.7 Code model card Evaluation Results table: 72.8. This source uses top_p=0.95; the existing Kimi K2.6 canonical setting uses top_p=1.0.
kimi-k2.7-code,mcpmark_verified,81.1,https://huggingface.co/moonshotai/Kimi-K2.7-Code,"{""context"":""262144 (256K)"",""effort"":""default"",""harness"":""official MCPMark; 100 steps; 32K max tokens per step"",""judge"":""programmatic verification on human-verified tasks"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""1.0; top_p=0.95"",""tools"":""benchmark-specified""}",True,model_card,verified,Kimi K2.7 Code model card Evaluation Results table: 81.1.
gpt-5.5,mcpmark_verified,92.9,https://huggingface.co/moonshotai/Kimi-K2.7-Code,"{""context"":""default"",""effort"":""xhigh"",""harness"":""official MCPMark; 100 steps; 32K max tokens per step"",""judge"":""programmatic verification on human-verified tasks"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,model_card,verified,"Kimi K2.7 Code model card Evaluation Results table: 92.9. Cross-model value reported by Moonshot AI. The source averages three MCPMark-Verified runs, while the GPT-5.5 canonical sampling is pass@1."
claude-opus-4.8,mcpmark_verified,76.4,https://huggingface.co/moonshotai/Kimi-K2.7-Code,"{""context"":""default"",""effort"":""xhigh"",""harness"":""official MCPMark; 100 steps; 32K max tokens per step"",""judge"":""programmatic verification on human-verified tasks"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,model_card,verified,"Kimi K2.7 Code model card Evaluation Results table: 76.4. Cross-model value reported by Moonshot AI. This source uses xhigh effort, or max for MLS-Bench-Lite; Anthropic's model default is high effort."
kimi-k2.7-code,mcpatlas,76.0,https://huggingface.co/moonshotai/Kimi-K2.7-Code,"{""context"":""262144 (256K)"",""effort"":""default"",""harness"":""official MCP-Atlas; 100 tool calls; 32K max tokens per step"",""judge"":""official MCP-Atlas claims-based judge"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""1.0; top_p=0.95"",""tools"":""benchmark-specified""}",False,model_card,verified,Kimi K2.7 Code model card Evaluation Results table: 76.0. The source uses 100 tool calls and 32K max tokens per step and does not identify whether the public 500-task subset was used; the current canonical row uses the public release and maxTurns=20.
claude-opus-4.8,mcpatlas,83.6,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""public 500; maxTurns=100; Gemini 3.1 Pro judge"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: MCP-Atlas = 83.6. Source setting: public 500; maxTurns=100; Gemini 3.1 Pro judge. Origin: Moonshot evaluation.
kimi-k3,gpqa_diamond,93.5,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""max/xhigh; no tools"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""none""}",True,tech_report,verified,Kimi K3 technical report: GPQA Diamond = 93.5. Source setting: max/xhigh; no tools. Origin: Moonshot evaluation.
kimi-k3,critpt,23.4,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""max/xhigh; no tools"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""none""}",True,tech_report,verified,Kimi K3 technical report: CritPt = 23.4. Source setting: max/xhigh; no tools. Origin: Artificial Analysis cited by Kimi report.
kimi-k3,aa_lcr,74.7,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""max/xhigh; no tools"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""none""}",True,tech_report,verified,Kimi K3 technical report: AA-LCR = 74.7. Source setting: max/xhigh; no tools. Origin: Artificial Analysis cited by Kimi report.
kimi-k3,hle,43.5,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""no tools; K3 max/temp=1/top-p=.95"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""none""}",True,tech_report,verified,Kimi K3 technical report: HLE-Full (no tools) = 43.5. Source setting: no tools; K3 max/temp=1/top-p=.95. Origin: Moonshot evaluation.
kimi-k3,hle_tools,56.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""general tools; pass@1"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""general tools""}",True,tech_report,verified,Kimi K3 technical report: HLE-Full (with tools) = 56.0. Source setting: general tools; pass@1. Origin: Moonshot evaluation.
kimi-k3,deep_swe_v1_1,67.5,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic coding harness per row"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: DeepSWE = 67.5. Source setting: agentic coding harness per row. Origin: Moonshot K3 run; competitors from DeepSWE leaderboard / GLM blog.
kimi-k3,program_bench,77.8,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic coding harness per row"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: ProgramBench = 77.8. Source setting: agentic coding harness per row. Origin: Moonshot K3 run; competitor values from Vals AI.
kimi-k3,terminal_bench_2_1,88.3,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic coding harness per row"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Terminal-Bench 2.1 = 88.3. Source setting: agentic coding harness per row. Origin: mixed official sources documented by Kimi report.
kimi-k3,frontier_swe,81.2,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic coding harness per row"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: FrontierSWE = 81.2. Source setting: agentic coding harness per row. Origin: Moonshot/OpenAI runs plus official leaderboard.
kimi-k3,swe_marathon_h20_2026_07_09,42.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""H20-calibrated pre-final-v1.1 branch (2026-07-09)"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: SWE-Marathon = 42.0. Source setting: H20-calibrated pre-final-v1.1 branch (2026-07-09). Origin: Moonshot evaluation.
kimi-k3,posttrain_bench,36.6,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""H20 avg@3 for K3/Fable/Sol; official H100 for Opus/GPT-5.5/GLM"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: PostTrainBench = 36.6. Source setting: H20 avg@3 for K3/Fable/Sol; official H100 for Opus/GPT-5.5/GLM. Origin: Moonshot H20 runs plus official H100 leaderboard values.
kimi-k3,mls_bench_lite,48.3,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic coding harness per row"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: MLS-Bench-Lite = 48.3. Source setting: agentic coding harness per row. Origin: Moonshot evaluation.
kimi-k3,scicode,58.7,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic coding harness per row"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: SciCode = 58.7. Source setting: agentic coding harness per row. Origin: Artificial Analysis cited by Kimi report.
kimi-k3,kimi_code_bench_v2,72.9,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic coding harness per row"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Kimi Code Bench 2.0 = 72.9. Source setting: agentic coding harness per row. Origin: Moonshot evaluation.
kimi-k3,browsecomp,91.2,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: BrowseComp = 91.2. Source setting: agentic benchmark harness. Origin: Moonshot K3 run; competitor provider sources.
kimi-k3,deepsearchqa_f1,95.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: DeepSearchQA (F1) = 95.0. Source setting: agentic benchmark harness. Origin: Moonshot evaluation.
kimi-k3,researchrubrics,76.2,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: ResearchRubrics = 76.2. Source setting: agentic benchmark harness. Origin: Moonshot evaluation.
kimi-k3,gdpval_aa_elo,1686.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: GDPval-AA v2 (Elo) = 1686.0. Source setting: agentic benchmark harness. Origin: Artificial Analysis cited by Kimi report.
kimi-k3,toolathlon_verified,76.5,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""Verified edition; official leaderboard snapshot"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Toolathlon-Verified = 76.5. Source setting: Verified edition; official leaderboard snapshot. Origin: official Toolathlon leaderboard cited by Kimi report.
kimi-k3,mcpmark_verified,94.5,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""verified tasks; sampling not restated in K3 report"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""not restated"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: MCPMark-Verified = 94.5. Source setting: verified tasks; sampling not restated in K3 report. Origin: Moonshot evaluation. Canonical compatibility remains under review; stored as best available primary with matches_canonical=false.
kimi-k3,mcpatlas,84.2,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""public 500; maxTurns=100; Gemini 3.1 Pro judge"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: MCP-Atlas = 84.2. Source setting: public 500; maxTurns=100; Gemini 3.1 Pro judge. Origin: Moonshot evaluation.
kimi-k3,automation_bench,30.8,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: AutomationBench = 30.8. Source setting: agentic benchmark harness. Origin: Moonshot evaluation.
kimi-k3,job_bench,54.3,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: JobBench = 54.3. Source setting: agentic benchmark harness. Origin: official JobBench leaderboard cited by Kimi report.
kimi-k3,aa_briefcase_elo,1548.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: AA-Briefcase (Elo) = 1548.0. Source setting: agentic benchmark harness. Origin: Artificial Analysis cited by Kimi report.
kimi-k3,agents_last_exam,28.3,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Agents' Last Exam = 28.3. Source setting: agentic benchmark harness. Origin: official leaderboard cited by Kimi report.
kimi-k3,apex_agents,41.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: APEX-Agents = 41.0. Source setting: agentic benchmark harness. Origin: Artificial Analysis cited by Kimi report.
kimi-k3,officeqa_pro,63.3,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""PDF corpus rendered as images; no machine-readable text"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: OfficeQA Pro = 63.3. Source setting: PDF corpus rendered as images; no machine-readable text. Origin: Moonshot evaluation.
kimi-k3,spreadsheetbench_2,34.8,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: SpreadsheetBench 2 = 34.8. Source setting: agentic benchmark harness. Origin: Moonshot evaluation.
kimi-k3,osworld_verified,84.8,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: OSWorld-Verified = 84.8. Source setting: agentic benchmark harness. Origin: Moonshot evaluation.
kimi-k3,osworld_2_0,58.3,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: OSWorld 2.0 = 58.3. Source setting: agentic benchmark harness. Origin: Moonshot evaluation.
kimi-k3,saas_bench,60.1,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: SaaS-Bench = 60.1. Source setting: agentic benchmark harness. Origin: Moonshot evaluation.
kimi-k3,tau3_banking,33.4,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: τ³-Banking = 33.4. Source setting: agentic benchmark harness. Origin: Artificial Analysis cited by Kimi report.
kimi-k3,harvey_lab_aa,94.6,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Harvey Lab-AA = 94.6. Source setting: agentic benchmark harness. Origin: Artificial Analysis cited by Kimi report.
kimi-k3,corpfin_v2,71.6,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: CorpFin v2 = 71.6. Source setting: agentic benchmark harness. Origin: Vals AI cited by Kimi report.
kimi-k3,finance_agent_v2,54.4,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Finance Agent v2 = 54.4. Source setting: agentic benchmark harness. Origin: Vals AI cited by Kimi report.
kimi-k3,legal_research_bench,44.2,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Legal Research Bench = 44.2. Source setting: agentic benchmark harness. Origin: Vals AI cited by Kimi report.
kimi-k3,worldvqa,51.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""forced-answer prompt; avg 3 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""none""}",False,tech_report,verified,Kimi K3 technical report: WorldVQA ForceAnswer = 51.0. Source setting: forced-answer prompt; avg 3 runs. Origin: Moonshot evaluation.
kimi-k3,omnidocbench,0.089,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""avg 3 runs; source 1-NED score normalized to NED"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: OmniDocBench = 0.089 NED. Source setting: avg 3 runs; source 1-NED score normalized to NED. Origin: Moonshot evaluation. The source reports 1-NED=91.1%; stored value is normalized NED=0.089.
kimi-k3,perception_bench,58.5,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""multimodal; avg 3 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: PerceptionBench = 58.5. Source setting: multimodal; avg 3 runs. Origin: Moonshot evaluation.
kimi-k3,video_mme,90.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""audio+visual+subtitles; avg 3 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""none""}",False,tech_report,verified,Kimi K3 technical report: Video-MME (w. sub) = 90.0. Source setting: audio+visual+subtitles; avg 3 runs. Origin: Moonshot evaluation.
kimi-k3,mmvu,82.1,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""avg 3 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: MMVU = 82.1. Source setting: avg 3 runs. Origin: Moonshot evaluation.
kimi-k3,babyvision,85.7,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""Python tool; avg 3 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""python""}",False,tech_report,verified,Kimi K3 technical report: BabyVision w/ python = 85.7. Source setting: Python tool; avg 3 runs. Origin: Moonshot evaluation.
kimi-k3,mmmu_pro,81.6,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""vision; avg 3 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""none""}",False,tech_report,verified,Kimi K3 technical report: MMMU-Pro (no Python) = 81.6. Source setting: vision; avg 3 runs. Origin: Moonshot evaluation.
kimi-k3,charxiv_reasoning,84.8,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""vision; avg 3 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""none""}",False,tech_report,verified,Kimi K3 technical report: CharXiv (RQ) (no Python) = 84.8. Source setting: vision; avg 3 runs. Origin: Moonshot evaluation.
kimi-k3,mathvision,94.3,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""vision; avg 3 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""none""}",False,tech_report,verified,Kimi K3 technical report: MathVision (no Python) = 94.3. Source setting: vision; avg 3 runs. Origin: Moonshot evaluation.
kimi-k3,zerobench_main,23.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""no Python; pass@5 / 5 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@5"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: ZeroBench (pass@5) (no Python) = 23.0. Source setting: no Python; pass@5 / 5 runs. Origin: Moonshot evaluation. Canonical compatibility remains under review; stored as best available primary with matches_canonical=false.
kimi-k3,zerobench_tools,41.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""Python tool; pass@5 / 5 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@5"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""python""}",False,tech_report,verified,Kimi K3 technical report: ZeroBench (pass@5) (with Python) = 41.0. Source setting: Python tool; pass@5 / 5 runs. Origin: Moonshot evaluation. Canonical compatibility remains under review; stored as best available primary with matches_canonical=false.
kimi-k3,coding_experience,56.6,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""provider-aligned harness: Kimi Code / Claude Code / Codex"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Coding Experience = 56.6. Source setting: provider-aligned harness: Kimi Code / Claude Code / Codex. Origin: Moonshot evaluation.
kimi-k3,clawbench_2_0,48.3,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""OpenClaw harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: 24/7 ClawBench 2.0 = 48.3. Source setting: OpenClaw harness. Origin: Moonshot evaluation.
kimi-k3,mira_bench,64.1,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""MIRA harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: MIRA Bench = 64.1. Source setting: MIRA harness. Origin: Moonshot evaluation.
kimi-k3,kaet,83.5,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""Kimi Code harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: KAET = 83.5. Source setting: Kimi Code harness. Origin: Moonshot evaluation.
kimi-k3,clif_bench,52.4,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""Kimi Code harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: CLIF Bench = 52.4. Source setting: Kimi Code harness. Origin: Moonshot evaluation.
kimi-k3,agentic_vision_bench,78.3,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""Kimi Code harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Agentic Vision Bench = 78.3. Source setting: Kimi Code harness. Origin: Moonshot evaluation.
kimi-k3,swarm_bench,76.3,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""Kimi Agent harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: SwarmBench = 76.3. Source setting: Kimi Agent harness. Origin: Moonshot evaluation.
kimi-k3,online_experience,77.9,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""Kimi Agent harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Online Experience = 77.9. Source setting: Kimi Agent harness. Origin: Moonshot evaluation.
kimi-k3,deepresearchbench,90.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""Kimi Agent harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: DeepResearchBench = 90.0. Source setting: Kimi Agent harness. Origin: Moonshot evaluation.
kimi-k3,finance_bench,62.6,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""N/A harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Finance Bench = 62.6. Source setting: N/A harness. Origin: Moonshot evaluation.
kimi-k3,kwv_bench,64.7,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""N/A harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: KWVBench = 64.7. Source setting: N/A harness. Origin: Moonshot evaluation.
kimi-k3,deck_bench,73.5,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""N/A harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: DECKBench = 73.5. Source setting: N/A harness. Origin: Moonshot evaluation.
kimi-k3,agent_behavior_bench,65.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""Kimi Work harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Agent Behavior Bench = 65.0. Source setting: Kimi Work harness. Origin: Moonshot evaluation.
kimi-k3,faithfulness,85.5,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""N/A harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Faithfulness (1 - hallucination rate) = 85.5. Source setting: N/A harness. Origin: Moonshot evaluation.
kimi-k3,chat_all_in_one_bench,85.2,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""Kimi Work harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Chat All-in-One Bench = 85.2. Source setting: Kimi Work harness. Origin: Moonshot evaluation.
kimi-k3,aa_intelligence_index_v4_1,57.1,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""live leaderboard snapshot as of 2026-07-23"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""leaderboard snapshot 2026-07-23"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: AA Intelligence Index v4.1 = 57.1. Source setting: live leaderboard snapshot as of 2026-07-23. Origin: third-party leaderboard cited by Kimi report.
kimi-k3,vals_index,74.7,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""live leaderboard snapshot as of 2026-07-23"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""leaderboard snapshot 2026-07-23"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Vals Index = 74.7. Source setting: live leaderboard snapshot as of 2026-07-23. Origin: third-party leaderboard cited by Kimi report.
kimi-k3,webdev_arena_elo,1678.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""live leaderboard snapshot as of 2026-07-23"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""leaderboard snapshot 2026-07-23"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: WebDevArena Elo = 1678. Source setting: live leaderboard snapshot as of 2026-07-23. Origin: third-party leaderboard cited by Kimi report.
kimi-k3,text_arena_elo,1486.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""live leaderboard snapshot as of 2026-07-23"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""leaderboard snapshot 2026-07-23"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: TextArena Elo = 1486. Source setting: live leaderboard snapshot as of 2026-07-23. Origin: third-party leaderboard cited by Kimi report.
kimi-k3,agent_arena,9.1,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""live leaderboard snapshot as of 2026-07-23"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""leaderboard snapshot 2026-07-23"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: AgentArena = 9.1. Source setting: live leaderboard snapshot as of 2026-07-23. Origin: third-party leaderboard cited by Kimi report.
kimi-k3,moonshot_exploit_development_suite,38.9,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""36-task in-house exploit suite"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Moonshot exploit-development suite = 38.9. Source setting: 36-task in-house exploit suite. Origin: Moonshot evaluation.
kimi-k3,exploitbench,32.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""independent joint assessment"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95 for reasoning/no-tool vision, top_p=1.0 for coding/agentic"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: ExploitBench (UK AISI / NIST CAISI) = 32.0. Source setting: independent joint assessment. Origin: UK AISI / NIST CAISI cited by Kimi report.
claude-fable-5,gpqa_diamond,92.6,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""max/xhigh; no tools"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Kimi K3 technical report: GPQA Diamond = 92.6. Source setting: max/xhigh; no tools. Origin: Moonshot evaluation.
claude-fable-5,critpt,28.6,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""max/xhigh; no tools"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Kimi K3 technical report: CritPt = 28.6. Source setting: max/xhigh; no tools. Origin: Artificial Analysis cited by Kimi report.
claude-fable-5,aa_lcr,70.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""max/xhigh; no tools"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Kimi K3 technical report: AA-LCR = 70.0. Source setting: max/xhigh; no tools. Origin: Artificial Analysis cited by Kimi report.
claude-fable-5,hle,56.5,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""benchmark-specific""}",True,model_card,verified,"Figure 8.10.1.B / Claude Fable 5 / max: 2,500 questions; 1M token cap; no tools and no context compaction."
claude-fable-5,hle_tools,63.9,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""benchmark-specific""}",True,model_card,verified,"Figure 8.10.1.A / Claude Fable 5 / max: 2,500 questions; 1M token cap; web/programmatic/code tools; no context compaction."
claude-fable-5,deep_swe_v1_1,70.0,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max, with fallback"",""harness"":""mini-swe-agent run by Datacurve"",""judge"":""isolated repository verifier"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic repository shell/editor""}",True,model_card,verified,"card46 DeepSWE v1.1 Pass@1 (%); source effort=max, with fallback; benchmark protocol matches canonical=True. Unknown source fields remain unknown."
claude-fable-5,program_bench,76.8,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic coding harness per row"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: ProgramBench = 76.8. Source setting: agentic coding harness per row. Origin: Moonshot K3 run; competitor values from Vals AI.
claude-fable-5,terminal_bench_2_1,84.3,https://media.x.ai/v1/website/4p5-5184fdf9.pdf,"{""context"":""unknown"",""effort"":""max, with fallback"",""harness"":""Grok Build for Grok; source-reported peer harnesses"",""judge"":""verified task-success evaluator"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""container terminal""}",True,model_card,verified,"card45 Terminal-Bench 2.1 Task success rate (%); source effort=max, with fallback; benchmark protocol matches canonical=True. Unknown source fields remain unknown."
claude-fable-5,frontier_swe,89.0,https://media.x.ai/v1/website/4p5-5184fdf9.pdf,"{""context"":""unknown"",""effort"":""max, with fallback"",""harness"":""public FrontierSWE harness; Grok CLI for Grok 4.5"",""judge"":""continuous partial-credit task scoring and dominance"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""mean@5"",""temperature"":""unknown"",""tools"":""agentic code execution""}",True,model_card,verified,"card45 FrontierSWE Dominance (%); source effort=max, with fallback; benchmark protocol matches canonical=True. Unknown source fields remain unknown."
claude-fable-5,swe_marathon_h20_2026_07_09,35.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""H20-calibrated pre-final-v1.1 branch (2026-07-09)"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: SWE-Marathon = 35.0. Source setting: H20-calibrated pre-final-v1.1 branch (2026-07-09). Origin: Moonshot evaluation.
claude-fable-5,posttrain_bench,41.4,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""H20 avg@3 for K3/Fable/Sol; official H100 for Opus/GPT-5.5/GLM"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: PostTrainBench = 41.4. Source setting: H20 avg@3 for K3/Fable/Sol; official H100 for Opus/GPT-5.5/GLM. Origin: Moonshot H20 runs plus official H100 leaderboard values.
claude-fable-5,mls_bench_lite,49.9,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic coding harness per row"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: MLS-Bench-Lite = 49.9. Source setting: agentic coding harness per row. Origin: Moonshot evaluation.
claude-fable-5,scicode,60.2,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic coding harness per row"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: SciCode = 60.2. Source setting: agentic coding harness per row. Origin: Artificial Analysis cited by Kimi report.
claude-fable-5,kimi_code_bench_v2,76.9,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic coding harness per row"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Kimi Code Bench 2.0 = 76.9. Source setting: agentic coding harness per row. Origin: Moonshot evaluation.
claude-fable-5,browsecomp,88.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: BrowseComp = 88.0. Source setting: agentic benchmark harness. Origin: Moonshot K3 run; competitor provider sources.
claude-fable-5,deepsearchqa_f1,94.7,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""benchmark-specific""}",True,model_card,verified,Figure 8.10.3.A / Claude Fable 5 / max: Mean F1 over 900 prompts in 17 fields; web tools; nominal 1M token budget.
claude-fable-5,gdpval_aa_elo,1747.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: GDPval-AA v2 (Elo) = 1747.0. Source setting: agentic benchmark harness. Origin: Artificial Analysis cited by Kimi report.
claude-fable-5,toolathlon_verified,77.9,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""Verified edition; official leaderboard snapshot"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Toolathlon-Verified = 77.9. Source setting: Verified edition; official leaderboard snapshot. Origin: official Toolathlon leaderboard cited by Kimi report.
claude-fable-5,mcpmark_verified,87.4,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""verified tasks; sampling not restated in K3 report"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""not restated"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: MCPMark-Verified = 87.4. Source setting: verified tasks; sampling not restated in K3 report. Origin: Moonshot evaluation. Canonical compatibility remains under review; stored as best available primary with matches_canonical=false.
claude-fable-5,mcpatlas,84.7,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""public 500; maxTurns=100; Gemini 3.1 Pro judge"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: MCP-Atlas = 84.7. Source setting: public 500; maxTurns=100; Gemini 3.1 Pro judge. Origin: Moonshot evaluation.
claude-fable-5,automation_bench,29.1,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: AutomationBench = 29.1. Source setting: agentic benchmark harness. Origin: Moonshot evaluation.
claude-fable-5,job_bench,57.4,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: JobBench = 57.4. Source setting: agentic benchmark harness. Origin: official JobBench leaderboard cited by Kimi report.
claude-fable-5,aa_briefcase_elo,1574.0,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max, with fallback"",""harness"":""Artificial Analysis AA-Briefcase harness"",""judge"":""rubric plus analytical and presentation pairwise Elo"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""offline multi-file knowledge-work tools""}",True,model_card,verified,"card46 AA Briefcase AA-Briefcase Elo; source effort=max, with fallback; benchmark protocol matches canonical=True. Unknown source fields remain unknown."
claude-fable-5,agents_last_exam,25.7,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: Agents' Last Exam = 25.7. Source setting: agentic benchmark harness. Origin: official leaderboard cited by Kimi report.
claude-fable-5,apex_agents,59.2,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max, with fallback"",""harness"":""Mercor APEX-Agents harness"",""judge"":""expert all-criteria-pass rubric"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""multi-application professional tools""}",True,model_card,verified,"card46 APEX-Agents Pass@1 (%); source effort=max, with fallback; benchmark protocol matches canonical=True. Unknown source fields remain unknown."
claude-fable-5,officeqa_pro,60.9,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max, with fallback"",""harness"":""OfficeQA Pro official harness"",""judge"":""OfficeQA Pro exact-match accuracy"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""document and rendered-image analysis""}",True,model_card,verified,"card46 OfficeQA Pro Accuracy (%); source effort=max, with fallback; benchmark protocol matches canonical=True. Unknown source fields remain unknown."
claude-fable-5,spreadsheetbench_2,34.7,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: SpreadsheetBench 2 = 34.7. Source setting: agentic benchmark harness. Origin: Moonshot evaluation.
claude-fable-5,osworld_verified,85.0,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""osworld"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""GUI computer-use harness""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.10.2.A, page 126: Claude Fable 5; OSWorld-Verified [summary]=85.0%. Source setting: effort=max; tools=GUI computer-use harness; sampling=avg 5 trials unless source states otherwise; harness=osworld."
claude-fable-5,osworld_2_0,66.1,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: OSWorld 2.0 = 66.1. Source setting: agentic benchmark harness. Origin: Moonshot evaluation.
claude-fable-5,tau3_banking,26.8,https://media.x.ai/v1/website/4p5-5184fdf9.pdf,"{""context"":""unknown"",""effort"":""max, with fallback"",""harness"":""Artificial Analysis tau3-banking harness"",""judge"":""final-state task accuracy"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""banking APIs and simulated user""}",True,model_card,verified,"card45 tau3-banking Accuracy (%); source effort=max, with fallback; benchmark protocol matches canonical=True. Unknown source fields remain unknown."
claude-fable-5,harvey_lab_aa,93.6,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Harvey Lab-AA = 93.6. Source setting: agentic benchmark harness. Origin: Artificial Analysis cited by Kimi report.
claude-fable-5,corpfin_v2,71.8,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: CorpFin v2 = 71.8. Source setting: agentic benchmark harness. Origin: Vals AI cited by Kimi report.
claude-fable-5,finance_agent_v2,56.3,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Finance Agent v2 = 56.3. Source setting: agentic benchmark harness. Origin: Vals AI cited by Kimi report.
claude-fable-5,legal_research_bench,49.5,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Legal Research Bench = 49.5. Source setting: agentic benchmark harness. Origin: Vals AI cited by Kimi report.
claude-fable-5,worldvqa,56.7,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""forced-answer prompt; avg 3 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,Kimi K3 technical report: WorldVQA ForceAnswer = 56.7. Source setting: forced-answer prompt; avg 3 runs. Origin: Moonshot evaluation.
claude-fable-5,omnidocbench,0.102,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""avg 3 runs; source 1-NED score normalized to NED"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: OmniDocBench = 0.102 NED. Source setting: avg 3 runs; source 1-NED score normalized to NED. Origin: Moonshot evaluation. The source reports 1-NED=89.8%; stored value is normalized NED=0.102.
claude-fable-5,perception_bench,57.2,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""multimodal; avg 3 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: PerceptionBench = 57.2. Source setting: multimodal; avg 3 runs. Origin: Moonshot evaluation.
claude-fable-5,babyvision,90.5,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""Python tool; avg 3 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""python""}",False,tech_report,verified,Kimi K3 technical report: BabyVision w/ python = 90.5. Source setting: Python tool; avg 3 runs. Origin: Moonshot evaluation.
claude-fable-5,mmmu_pro,81.2,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""vision; avg 3 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,Kimi K3 technical report: MMMU-Pro (no Python) = 81.2. Source setting: vision; avg 3 runs. Origin: Moonshot evaluation.
claude-fable-5,charxiv_reasoning,88.9,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""vision; avg 3 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,Kimi K3 technical report: CharXiv (RQ) (no Python) = 88.9. Source setting: vision; avg 3 runs. Origin: Moonshot evaluation.
claude-fable-5,mathvision,94.8,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""vision; avg 3 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,Kimi K3 technical report: MathVision (no Python) = 94.8. Source setting: vision; avg 3 runs. Origin: Moonshot evaluation.
claude-fable-5,zerobench_main,23.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""no Python; pass@5 / 5 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""pass@5"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: ZeroBench (pass@5) (no Python) = 23.0. Source setting: no Python; pass@5 / 5 runs. Origin: Moonshot evaluation. Canonical compatibility remains under review; stored as best available primary with matches_canonical=false.
claude-fable-5,zerobench_tools,46.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""Python tool; pass@5 / 5 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""pass@5"",""temperature"":""default"",""tools"":""python""}",False,tech_report,verified,Kimi K3 technical report: ZeroBench (pass@5) (with Python) = 46.0. Source setting: Python tool; pass@5 / 5 runs. Origin: Moonshot evaluation. Canonical compatibility remains under review; stored as best available primary with matches_canonical=false.
claude-fable-5,coding_experience,59.8,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""provider-aligned harness: Kimi Code / Claude Code / Codex"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Coding Experience = 59.8. Source setting: provider-aligned harness: Kimi Code / Claude Code / Codex. Origin: Moonshot evaluation.
claude-fable-5,clawbench_2_0,47.4,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""OpenClaw harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: 24/7 ClawBench 2.0 = 47.4. Source setting: OpenClaw harness. Origin: Moonshot evaluation.
claude-fable-5,mira_bench,72.9,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""MIRA harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: MIRA Bench = 72.9. Source setting: MIRA harness. Origin: Moonshot evaluation.
claude-fable-5,agentic_vision_bench,81.1,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""Kimi Code harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Agentic Vision Bench = 81.1. Source setting: Kimi Code harness. Origin: Moonshot evaluation.
claude-fable-5,online_experience,74.2,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""Kimi Agent harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Online Experience = 74.2. Source setting: Kimi Agent harness. Origin: Moonshot evaluation.
claude-fable-5,kwv_bench,63.6,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""N/A harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: KWVBench = 63.6. Source setting: N/A harness. Origin: Moonshot evaluation.
claude-fable-5,deck_bench,73.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""N/A harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: DECKBench = 73.0. Source setting: N/A harness. Origin: Moonshot evaluation.
claude-fable-5,agent_behavior_bench,75.5,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""Kimi Work harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Agent Behavior Bench = 75.5. Source setting: Kimi Work harness. Origin: Moonshot evaluation.
claude-fable-5,chat_all_in_one_bench,88.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""Kimi Work harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Chat All-in-One Bench = 88.0. Source setting: Kimi Work harness. Origin: Moonshot evaluation.
claude-fable-5,aa_intelligence_index_v4_1,59.9,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""live leaderboard snapshot as of 2026-07-23"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""leaderboard snapshot 2026-07-23"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: AA Intelligence Index v4.1 = 59.9. Source setting: live leaderboard snapshot as of 2026-07-23. Origin: third-party leaderboard cited by Kimi report.
claude-fable-5,vals_index,75.1,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max, with fallback"",""harness"":""Vals Index live harness"",""judge"":""Vals weighted suite score"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""benchmark-specified professional tools""}",True,model_card,verified,"card46 Vals Index Suite score (%); source effort=max, with fallback; benchmark protocol matches canonical=True. Unknown source fields remain unknown."
claude-fable-5,webdev_arena_elo,1634.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""live leaderboard snapshot as of 2026-07-23"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""leaderboard snapshot 2026-07-23"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: WebDevArena Elo = 1634. Source setting: live leaderboard snapshot as of 2026-07-23. Origin: third-party leaderboard cited by Kimi report.
claude-fable-5,text_arena_elo,1507.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""live leaderboard snapshot as of 2026-07-23"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""leaderboard snapshot 2026-07-23"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: TextArena Elo = 1507. Source setting: live leaderboard snapshot as of 2026-07-23. Origin: third-party leaderboard cited by Kimi report.
claude-fable-5,agent_arena,12.7,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""live leaderboard snapshot as of 2026-07-23"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default; product fallback enabled"",""sampling"":""leaderboard snapshot 2026-07-23"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: AgentArena = 12.7. Source setting: live leaderboard snapshot as of 2026-07-23. Origin: third-party leaderboard cited by Kimi report.
gpt-5.6-sol,gpqa_diamond,94.1,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""max/xhigh; no tools"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Kimi K3 technical report: GPQA Diamond = 94.1. Source setting: max/xhigh; no tools. Origin: Moonshot evaluation.
gpt-5.6-sol,critpt,32.3,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""max/xhigh; no tools"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Kimi K3 technical report: CritPt = 32.3. Source setting: max/xhigh; no tools. Origin: Artificial Analysis cited by Kimi report.
gpt-5.6-sol,aa_lcr,73.7,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""max/xhigh; no tools"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Kimi K3 technical report: AA-LCR = 73.7. Source setting: max/xhigh; no tools. Origin: Artificial Analysis cited by Kimi report.
gpt-5.6-sol,hle,44.5,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""no tools; K3 max/temp=1/top-p=.95"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Kimi K3 technical report: HLE-Full (no tools) = 44.5. Source setting: no tools; K3 max/temp=1/top-p=.95. Origin: Moonshot evaluation.
gpt-5.6-sol,hle_tools,58.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""general tools; pass@1"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""general tools""}",True,tech_report,verified,Kimi K3 technical report: HLE-Full (with tools) = 58.0. Source setting: general tools; pass@1. Origin: Moonshot evaluation.
gpt-5.6-sol,deep_swe_v1_1,73.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic coding harness per row"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: DeepSWE = 73.0. Source setting: agentic coding harness per row. Origin: Moonshot K3 run; competitors from DeepSWE leaderboard / GLM blog.
gpt-5.6-sol,program_bench,77.6,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic coding harness per row"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: ProgramBench = 77.6. Source setting: agentic coding harness per row. Origin: Moonshot K3 run; competitor values from Vals AI.
gpt-5.6-sol,terminal_bench_2_1,88.8,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic coding harness per row"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Terminal-Bench 2.1 = 88.8. Source setting: agentic coding harness per row. Origin: mixed official sources documented by Kimi report.
gpt-5.6-sol,frontier_swe,71.3,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic coding harness per row"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: FrontierSWE = 71.3. Source setting: agentic coding harness per row. Origin: Moonshot/OpenAI runs plus official leaderboard.
gpt-5.6-sol,swe_marathon_h20_2026_07_09,39.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""H20-calibrated pre-final-v1.1 branch (2026-07-09)"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: SWE-Marathon = 39.0. Source setting: H20-calibrated pre-final-v1.1 branch (2026-07-09). Origin: Moonshot evaluation.
gpt-5.6-sol,posttrain_bench,34.6,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""H20 avg@3 for K3/Fable/Sol; official H100 for Opus/GPT-5.5/GLM"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: PostTrainBench = 34.6. Source setting: H20 avg@3 for K3/Fable/Sol; official H100 for Opus/GPT-5.5/GLM. Origin: Moonshot H20 runs plus official H100 leaderboard values.
gpt-5.6-sol,mls_bench_lite,46.2,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic coding harness per row"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: MLS-Bench-Lite = 46.2. Source setting: agentic coding harness per row. Origin: Moonshot evaluation.
gpt-5.6-sol,scicode,56.1,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic coding harness per row"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: SciCode = 56.1. Source setting: agentic coding harness per row. Origin: Artificial Analysis cited by Kimi report.
gpt-5.6-sol,kimi_code_bench_v2,64.8,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic coding harness per row"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Kimi Code Bench 2.0 = 64.8. Source setting: agentic coding harness per row. Origin: Moonshot evaluation.
gpt-5.6-sol,browsecomp,90.4,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: BrowseComp = 90.4. Source setting: agentic benchmark harness. Origin: Moonshot K3 run; competitor provider sources.
gpt-5.6-sol,researchrubrics,73.8,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: ResearchRubrics = 73.8. Source setting: agentic benchmark harness. Origin: Moonshot evaluation.
gpt-5.6-sol,gdpval_aa_elo,1736.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: GDPval-AA v2 (Elo) = 1736.0. Source setting: agentic benchmark harness. Origin: Artificial Analysis cited by Kimi report.
gpt-5.6-sol,toolathlon_verified,74.9,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""Verified edition; official leaderboard snapshot"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Toolathlon-Verified = 74.9. Source setting: Verified edition; official leaderboard snapshot. Origin: official Toolathlon leaderboard cited by Kimi report.
gpt-5.6-sol,mcpmark_verified,92.9,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""verified tasks; sampling not restated in K3 report"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""not restated"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: MCPMark-Verified = 92.9. Source setting: verified tasks; sampling not restated in K3 report. Origin: Moonshot evaluation. Canonical compatibility remains under review; stored as best available primary with matches_canonical=false.
gpt-5.6-sol,mcpatlas,83.6,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""public 500; maxTurns=100; Gemini 3.1 Pro judge"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: MCP-Atlas = 83.6. Source setting: public 500; maxTurns=100; Gemini 3.1 Pro judge. Origin: Moonshot evaluation.
gpt-5.6-sol,automation_bench,29.7,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: AutomationBench = 29.7. Source setting: agentic benchmark harness. Origin: Moonshot evaluation.
gpt-5.6-sol,job_bench,45.4,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: JobBench = 45.4. Source setting: agentic benchmark harness. Origin: official JobBench leaderboard cited by Kimi report.
gpt-5.6-sol,aa_briefcase_elo,1495.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: AA-Briefcase (Elo) = 1495.0. Source setting: agentic benchmark harness. Origin: Artificial Analysis cited by Kimi report.
gpt-5.6-sol,agents_last_exam,29.6,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Agents' Last Exam = 29.6. Source setting: agentic benchmark harness. Origin: official leaderboard cited by Kimi report.
gpt-5.6-sol,apex_agents,39.9,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: APEX-Agents = 39.9. Source setting: agentic benchmark harness. Origin: Artificial Analysis cited by Kimi report.
gpt-5.6-sol,officeqa_pro,63.2,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""PDF corpus rendered as images; no machine-readable text"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: OfficeQA Pro = 63.2. Source setting: PDF corpus rendered as images; no machine-readable text. Origin: Moonshot evaluation.
gpt-5.6-sol,spreadsheetbench_2,32.4,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: SpreadsheetBench 2 = 32.4. Source setting: agentic benchmark harness. Origin: Moonshot evaluation.
gpt-5.6-sol,osworld_verified,83.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: OSWorld-Verified = 83.0. Source setting: agentic benchmark harness. Origin: Moonshot evaluation.
gpt-5.6-sol,osworld_2_0,62.6,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: OSWorld 2.0 = 62.6. Source setting: agentic benchmark harness. Origin: Moonshot evaluation.
gpt-5.6-sol,saas_bench,61.4,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: SaaS-Bench = 61.4. Source setting: agentic benchmark harness. Origin: Moonshot evaluation.
gpt-5.6-sol,tau3_banking,33.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: τ³-Banking = 33.0. Source setting: agentic benchmark harness. Origin: Artificial Analysis cited by Kimi report.
gpt-5.6-sol,harvey_lab_aa,87.2,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Harvey Lab-AA = 87.2. Source setting: agentic benchmark harness. Origin: Artificial Analysis cited by Kimi report.
gpt-5.6-sol,corpfin_v2,64.4,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: CorpFin v2 = 64.4. Source setting: agentic benchmark harness. Origin: Vals AI cited by Kimi report.
gpt-5.6-sol,finance_agent_v2,53.8,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Finance Agent v2 = 53.8. Source setting: agentic benchmark harness. Origin: Vals AI cited by Kimi report.
gpt-5.6-sol,legal_research_bench,48.1,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Legal Research Bench = 48.1. Source setting: agentic benchmark harness. Origin: Vals AI cited by Kimi report.
gpt-5.6-sol,worldvqa,41.8,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""forced-answer prompt; avg 3 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,Kimi K3 technical report: WorldVQA ForceAnswer = 41.8. Source setting: forced-answer prompt; avg 3 runs. Origin: Moonshot evaluation.
gpt-5.6-sol,omnidocbench,0.142,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""avg 3 runs; source 1-NED score normalized to NED"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: OmniDocBench = 0.142 NED. Source setting: avg 3 runs; source 1-NED score normalized to NED. Origin: Moonshot evaluation. The source reports 1-NED=85.8%; stored value is normalized NED=0.142.
gpt-5.6-sol,perception_bench,59.7,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""multimodal; avg 3 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: PerceptionBench = 59.7. Source setting: multimodal; avg 3 runs. Origin: Moonshot evaluation.
gpt-5.6-sol,video_mme,89.5,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""audio+visual+subtitles; avg 3 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,Kimi K3 technical report: Video-MME (w. sub) = 89.5. Source setting: audio+visual+subtitles; avg 3 runs. Origin: Moonshot evaluation.
gpt-5.6-sol,mmvu,81.2,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""avg 3 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: MMVU = 81.2. Source setting: avg 3 runs. Origin: Moonshot evaluation.
gpt-5.6-sol,babyvision,88.9,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""Python tool; avg 3 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""python""}",False,tech_report,verified,Kimi K3 technical report: BabyVision w/ python = 88.9. Source setting: Python tool; avg 3 runs. Origin: Moonshot evaluation.
gpt-5.6-sol,mmmu_pro,83.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""vision; avg 3 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,Kimi K3 technical report: MMMU-Pro (no Python) = 83.0. Source setting: vision; avg 3 runs. Origin: Moonshot evaluation.
gpt-5.6-sol,charxiv_reasoning,84.6,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""vision; avg 3 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,Kimi K3 technical report: CharXiv (RQ) (no Python) = 84.6. Source setting: vision; avg 3 runs. Origin: Moonshot evaluation.
gpt-5.6-sol,mathvision,95.8,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""vision; avg 3 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,Kimi K3 technical report: MathVision (no Python) = 95.8. Source setting: vision; avg 3 runs. Origin: Moonshot evaluation.
gpt-5.6-sol,zerobench_main,17.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""no Python; pass@5 / 5 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@5"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: ZeroBench (pass@5) (no Python) = 17.0. Source setting: no Python; pass@5 / 5 runs. Origin: Moonshot evaluation. Canonical compatibility remains under review; stored as best available primary with matches_canonical=false.
gpt-5.6-sol,zerobench_tools,35.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""Python tool; pass@5 / 5 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@5"",""temperature"":""default"",""tools"":""python""}",False,tech_report,verified,Kimi K3 technical report: ZeroBench (pass@5) (with Python) = 35.0. Source setting: Python tool; pass@5 / 5 runs. Origin: Moonshot evaluation. Canonical compatibility remains under review; stored as best available primary with matches_canonical=false.
gpt-5.6-sol,coding_experience,59.3,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""provider-aligned harness: Kimi Code / Claude Code / Codex"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Coding Experience = 59.3. Source setting: provider-aligned harness: Kimi Code / Claude Code / Codex. Origin: Moonshot evaluation.
gpt-5.6-sol,clawbench_2_0,52.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""OpenClaw harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: 24/7 ClawBench 2.0 = 52.0. Source setting: OpenClaw harness. Origin: Moonshot evaluation.
gpt-5.6-sol,mira_bench,62.2,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""MIRA harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: MIRA Bench = 62.2. Source setting: MIRA harness. Origin: Moonshot evaluation.
gpt-5.6-sol,kaet,85.4,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""Kimi Code harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: KAET = 85.4. Source setting: Kimi Code harness. Origin: Moonshot evaluation.
gpt-5.6-sol,clif_bench,50.6,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""Kimi Code harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: CLIF Bench = 50.6. Source setting: Kimi Code harness. Origin: Moonshot evaluation.
gpt-5.6-sol,agentic_vision_bench,82.9,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""Kimi Code harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Agentic Vision Bench = 82.9. Source setting: Kimi Code harness. Origin: Moonshot evaluation.
gpt-5.6-sol,swarm_bench,73.2,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""Kimi Agent harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: SwarmBench = 73.2. Source setting: Kimi Agent harness. Origin: Moonshot evaluation.
gpt-5.6-sol,online_experience,84.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""Kimi Agent harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Online Experience = 84.0. Source setting: Kimi Agent harness. Origin: Moonshot evaluation.
gpt-5.6-sol,deepresearchbench,85.3,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""Kimi Agent harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: DeepResearchBench = 85.3. Source setting: Kimi Agent harness. Origin: Moonshot evaluation.
gpt-5.6-sol,finance_bench,62.7,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""N/A harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Finance Bench = 62.7. Source setting: N/A harness. Origin: Moonshot evaluation.
gpt-5.6-sol,kwv_bench,66.9,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""N/A harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: KWVBench = 66.9. Source setting: N/A harness. Origin: Moonshot evaluation.
gpt-5.6-sol,deck_bench,74.7,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""N/A harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: DECKBench = 74.7. Source setting: N/A harness. Origin: Moonshot evaluation.
gpt-5.6-sol,agent_behavior_bench,76.4,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""Kimi Work harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Agent Behavior Bench = 76.4. Source setting: Kimi Work harness. Origin: Moonshot evaluation.
gpt-5.6-sol,faithfulness,84.8,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""N/A harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Faithfulness (1 - hallucination rate) = 84.8. Source setting: N/A harness. Origin: Moonshot evaluation.
gpt-5.6-sol,chat_all_in_one_bench,79.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""Kimi Work harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Chat All-in-One Bench = 79.0. Source setting: Kimi Work harness. Origin: Moonshot evaluation.
gpt-5.6-sol,aa_intelligence_index_v4_1,58.9,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""live leaderboard snapshot as of 2026-07-23"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""leaderboard snapshot 2026-07-23"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: AA Intelligence Index v4.1 = 58.9. Source setting: live leaderboard snapshot as of 2026-07-23. Origin: third-party leaderboard cited by Kimi report.
gpt-5.6-sol,vals_index,73.1,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""live leaderboard snapshot as of 2026-07-23"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""leaderboard snapshot 2026-07-23"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Vals Index = 73.1. Source setting: live leaderboard snapshot as of 2026-07-23. Origin: third-party leaderboard cited by Kimi report.
gpt-5.6-sol,webdev_arena_elo,1630.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""live leaderboard snapshot as of 2026-07-23"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""leaderboard snapshot 2026-07-23"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: WebDevArena Elo = 1630. Source setting: live leaderboard snapshot as of 2026-07-23. Origin: third-party leaderboard cited by Kimi report.
gpt-5.6-sol,text_arena_elo,1485.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""live leaderboard snapshot as of 2026-07-23"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""leaderboard snapshot 2026-07-23"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: TextArena Elo = 1485. Source setting: live leaderboard snapshot as of 2026-07-23. Origin: third-party leaderboard cited by Kimi report.
gpt-5.6-sol,agent_arena,10.1,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""live leaderboard snapshot as of 2026-07-23"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""leaderboard snapshot 2026-07-23"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: AgentArena = 10.1. Source setting: live leaderboard snapshot as of 2026-07-23. Origin: third-party leaderboard cited by Kimi report.
claude-opus-4.8,gpqa_diamond,91.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""max/xhigh; no tools"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,Kimi K3 technical report: GPQA Diamond = 91.0. Source setting: max/xhigh; no tools. Origin: Moonshot evaluation.
claude-opus-4.8,critpt,20.9,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""max/xhigh; no tools"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,Kimi K3 technical report: CritPt = 20.9. Source setting: max/xhigh; no tools. Origin: Artificial Analysis cited by Kimi report.
claude-opus-4.8,aa_lcr,67.7,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""max/xhigh; no tools"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,Kimi K3 technical report: AA-LCR = 67.7. Source setting: max/xhigh; no tools. Origin: Artificial Analysis cited by Kimi report.
claude-opus-4.8,hle,47.1,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""high"",""tools"":""benchmark-specific""}",True,model_card,verified,"Figure 8.10.1.B / Claude Opus 4.8 / high: 2,500 questions; 1M token cap; no tools and no context compaction."
claude-opus-4.8,hle_tools,55.7,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""high"",""harness"":""hle"",""sampling"":""pass@1"",""tools"":""web search, fetch, and code execution""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.9.1.B, page 123: Claude Opus 4.8; Humanity’s Last Exam [high]=55.7%. Source setting: effort=high; tools=web search, fetch, and code execution; sampling=pass@1; harness=hle."
claude-opus-4.8,deep_swe_v1_1,51.8,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""high"",""tools"":""benchmark-specific""}",True,model_card,verified,Figure 8.3.A / Claude Opus 4.8 / high: 113 tasks; five trials.
claude-opus-4.8,terminal_bench_2_1,84.6,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic coding harness per row"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: Terminal-Bench 2.1 = 84.6. Source setting: agentic coding harness per row. Origin: mixed official sources documented by Kimi report.
claude-opus-4.8,frontier_swe,66.7,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic coding harness per row"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: FrontierSWE = 66.7. Source setting: agentic coding harness per row. Origin: Moonshot/OpenAI runs plus official leaderboard.
claude-opus-4.8,swe_marathon_h20_2026_07_09,40.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""H20-calibrated pre-final-v1.1 branch (2026-07-09)"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: SWE-Marathon = 40.0. Source setting: H20-calibrated pre-final-v1.1 branch (2026-07-09). Origin: Moonshot evaluation.
claude-opus-4.8,posttrain_bench,34.1,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""H20 avg@3 for K3/Fable/Sol; official H100 for Opus/GPT-5.5/GLM"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: PostTrainBench = 34.1. Source setting: H20 avg@3 for K3/Fable/Sol; official H100 for Opus/GPT-5.5/GLM. Origin: Moonshot H20 runs plus official H100 leaderboard values.
claude-opus-4.8,scicode,53.5,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic coding harness per row"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: SciCode = 53.5. Source setting: agentic coding harness per row. Origin: Artificial Analysis cited by Kimi report.
claude-opus-4.8,browsecomp,82.0,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""high"",""tools"":""benchmark-specific""}",True,model_card,verified,Figure 8.10.2.B / Claude Opus 4.8 / high: Fixed 10M token budget with web/programmatic/code tools and context compaction.
claude-opus-4.8,deepsearchqa_f1,91.0,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""high"",""tools"":""benchmark-specific""}",True,model_card,verified,Figure 8.10.3.A / Claude Opus 4.8 / high: Mean F1 over 900 prompts in 17 fields; web tools; nominal 1M token budget.
claude-opus-4.8,researchrubrics,73.5,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: ResearchRubrics = 73.5. Source setting: agentic benchmark harness. Origin: Moonshot evaluation.
claude-opus-4.8,gdpval_aa_elo,1593.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: GDPval-AA v2 (Elo) = 1593.0. Source setting: agentic benchmark harness. Origin: Artificial Analysis cited by Kimi report.
claude-opus-4.8,toolathlon_verified,76.2,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""Verified edition; official leaderboard snapshot"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: Toolathlon-Verified = 76.2. Source setting: Verified edition; official leaderboard snapshot. Origin: official Toolathlon leaderboard cited by Kimi report.
claude-opus-4.8,automation_bench,27.2,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: AutomationBench = 27.2. Source setting: agentic benchmark harness. Origin: Moonshot evaluation.
claude-opus-4.8,job_bench,48.4,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: JobBench = 48.4. Source setting: agentic benchmark harness. Origin: official JobBench leaderboard cited by Kimi report.
claude-opus-4.8,aa_briefcase_elo,1354.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: AA-Briefcase (Elo) = 1354.0. Source setting: agentic benchmark harness. Origin: Artificial Analysis cited by Kimi report.
claude-opus-4.8,agents_last_exam,27.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: Agents' Last Exam = 27.0. Source setting: agentic benchmark harness. Origin: official leaderboard cited by Kimi report.
claude-opus-4.8,apex_agents,39.4,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: APEX-Agents = 39.4. Source setting: agentic benchmark harness. Origin: Artificial Analysis cited by Kimi report.
claude-opus-4.8,officeqa_pro,63.9,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""PDF corpus rendered as images; no machine-readable text"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: OfficeQA Pro = 63.9. Source setting: PDF corpus rendered as images; no machine-readable text. Origin: Moonshot evaluation.
claude-opus-4.8,spreadsheetbench_2,31.6,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: SpreadsheetBench 2 = 31.6. Source setting: agentic benchmark harness. Origin: Moonshot evaluation.
claude-opus-4.8,osworld_verified,83.4,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""osworld"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""GUI computer-use harness""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.10.2.A, page 126: Claude Opus 4.8; OSWorld-Verified [summary]=83.4%. Source setting: effort=max; tools=GUI computer-use harness; sampling=avg 5 trials unless source states otherwise; harness=osworld."
claude-opus-4.8,osworld_2_0,55.7,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: OSWorld 2.0 = 55.7. Source setting: agentic benchmark harness. Origin: Moonshot evaluation.
claude-opus-4.8,saas_bench,56.1,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: SaaS-Bench = 56.1. Source setting: agentic benchmark harness. Origin: Moonshot evaluation.
claude-opus-4.8,tau3_banking,27.6,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: τ³-Banking = 27.6. Source setting: agentic benchmark harness. Origin: Artificial Analysis cited by Kimi report.
claude-opus-4.8,harvey_lab_aa,91.1,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: Harvey Lab-AA = 91.1. Source setting: agentic benchmark harness. Origin: Artificial Analysis cited by Kimi report.
claude-opus-4.8,corpfin_v2,66.7,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: CorpFin v2 = 66.7. Source setting: agentic benchmark harness. Origin: Vals AI cited by Kimi report.
claude-opus-4.8,finance_agent_v2,53.9,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: Finance Agent v2 = 53.9. Source setting: agentic benchmark harness. Origin: Vals AI cited by Kimi report.
claude-opus-4.8,legal_research_bench,43.8,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: Legal Research Bench = 43.8. Source setting: agentic benchmark harness. Origin: Vals AI cited by Kimi report.
claude-opus-4.8,worldvqa,39.1,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""forced-answer prompt; avg 3 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,Kimi K3 technical report: WorldVQA ForceAnswer = 39.1. Source setting: forced-answer prompt; avg 3 runs. Origin: Moonshot evaluation.
claude-opus-4.8,omnidocbench,0.121,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""avg 3 runs; source 1-NED score normalized to NED"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: OmniDocBench = 0.121 NED. Source setting: avg 3 runs; source 1-NED score normalized to NED. Origin: Moonshot evaluation. The source reports 1-NED=87.9%; stored value is normalized NED=0.121.
claude-opus-4.8,perception_bench,47.2,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""multimodal; avg 3 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: PerceptionBench = 47.2. Source setting: multimodal; avg 3 runs. Origin: Moonshot evaluation.
claude-opus-4.8,video_mme,86.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""audio+visual+subtitles; avg 3 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,Kimi K3 technical report: Video-MME (w. sub) = 86.0. Source setting: audio+visual+subtitles; avg 3 runs. Origin: Moonshot evaluation.
claude-opus-4.8,mmvu,79.2,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""avg 3 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: MMVU = 79.2. Source setting: avg 3 runs. Origin: Moonshot evaluation.
claude-opus-4.8,babyvision,81.2,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""Python tool; avg 3 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""python""}",False,tech_report,verified,Kimi K3 technical report: BabyVision w/ python = 81.2. Source setting: Python tool; avg 3 runs. Origin: Moonshot evaluation.
claude-opus-4.8,mmmu_pro,78.9,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""vision; avg 3 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,Kimi K3 technical report: MMMU-Pro (no Python) = 78.9. Source setting: vision; avg 3 runs. Origin: Moonshot evaluation.
claude-opus-4.8,charxiv_reasoning,80.5,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""vision"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.10.5.A, page 130: Claude Opus 4.8; CharXiv Reasoning [no tools]=80.5. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=vision."
claude-opus-4.8,mathvision,86.7,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""vision; avg 3 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,Kimi K3 technical report: MathVision (no Python) = 86.7. Source setting: vision; avg 3 runs. Origin: Moonshot evaluation.
claude-opus-4.8,zerobench_main,17.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""no Python; pass@5 / 5 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@5"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: ZeroBench (pass@5) (no Python) = 17.0. Source setting: no Python; pass@5 / 5 runs. Origin: Moonshot evaluation.
claude-opus-4.8,zerobench_tools,34.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""Python tool; pass@5 / 5 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@5"",""temperature"":""default"",""tools"":""python""}",False,tech_report,verified,Kimi K3 technical report: ZeroBench (pass@5) (with Python) = 34.0. Source setting: Python tool; pass@5 / 5 runs. Origin: Moonshot evaluation.
claude-opus-4.8,coding_experience,58.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""provider-aligned harness: Kimi Code / Claude Code / Codex"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: Coding Experience = 58.0. Source setting: provider-aligned harness: Kimi Code / Claude Code / Codex. Origin: Moonshot evaluation.
claude-opus-4.8,clawbench_2_0,47.2,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""OpenClaw harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: 24/7 ClawBench 2.0 = 47.2. Source setting: OpenClaw harness. Origin: Moonshot evaluation.
claude-opus-4.8,mira_bench,59.8,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""MIRA harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: MIRA Bench = 59.8. Source setting: MIRA harness. Origin: Moonshot evaluation.
claude-opus-4.8,kaet,78.7,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""Kimi Code harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: KAET = 78.7. Source setting: Kimi Code harness. Origin: Moonshot evaluation.
claude-opus-4.8,clif_bench,48.8,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""Kimi Code harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: CLIF Bench = 48.8. Source setting: Kimi Code harness. Origin: Moonshot evaluation.
claude-opus-4.8,agentic_vision_bench,82.8,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""Kimi Code harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: Agentic Vision Bench = 82.8. Source setting: Kimi Code harness. Origin: Moonshot evaluation.
claude-opus-4.8,swarm_bench,72.6,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""Kimi Agent harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: SwarmBench = 72.6. Source setting: Kimi Agent harness. Origin: Moonshot evaluation.
claude-opus-4.8,online_experience,69.4,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""Kimi Agent harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: Online Experience = 69.4. Source setting: Kimi Agent harness. Origin: Moonshot evaluation.
claude-opus-4.8,deepresearchbench,87.2,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""Kimi Agent harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: DeepResearchBench = 87.2. Source setting: Kimi Agent harness. Origin: Moonshot evaluation.
claude-opus-4.8,finance_bench,60.7,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""N/A harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: Finance Bench = 60.7. Source setting: N/A harness. Origin: Moonshot evaluation.
claude-opus-4.8,kwv_bench,61.7,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""N/A harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: KWVBench = 61.7. Source setting: N/A harness. Origin: Moonshot evaluation.
claude-opus-4.8,deck_bench,66.9,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""N/A harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: DECKBench = 66.9. Source setting: N/A harness. Origin: Moonshot evaluation.
claude-opus-4.8,agent_behavior_bench,65.7,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""Kimi Work harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: Agent Behavior Bench = 65.7. Source setting: Kimi Work harness. Origin: Moonshot evaluation.
claude-opus-4.8,faithfulness,83.6,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""N/A harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: Faithfulness (1 - hallucination rate) = 83.6. Source setting: N/A harness. Origin: Moonshot evaluation.
claude-opus-4.8,chat_all_in_one_bench,83.8,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""Kimi Work harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: Chat All-in-One Bench = 83.8. Source setting: Kimi Work harness. Origin: Moonshot evaluation.
claude-opus-4.8,aa_intelligence_index_v4_1,55.7,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""live leaderboard snapshot as of 2026-07-23"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""leaderboard snapshot 2026-07-23"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: AA Intelligence Index v4.1 = 55.7. Source setting: live leaderboard snapshot as of 2026-07-23. Origin: third-party leaderboard cited by Kimi report.
claude-opus-4.8,vals_index,70.4,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""live leaderboard snapshot as of 2026-07-23"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""leaderboard snapshot 2026-07-23"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: Vals Index = 70.4. Source setting: live leaderboard snapshot as of 2026-07-23. Origin: third-party leaderboard cited by Kimi report.
claude-opus-4.8,webdev_arena_elo,1565.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""live leaderboard snapshot as of 2026-07-23"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""leaderboard snapshot 2026-07-23"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: WebDevArena Elo = 1565. Source setting: live leaderboard snapshot as of 2026-07-23. Origin: third-party leaderboard cited by Kimi report.
claude-opus-4.8,text_arena_elo,1484.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""high"",""harness"":""live leaderboard snapshot as of 2026-07-23"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""leaderboard snapshot 2026-07-23"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: TextArena Elo = 1484. Source setting: live leaderboard snapshot as of 2026-07-23. Origin: third-party leaderboard cited by Kimi report.
claude-opus-4.8,agent_arena,9.8,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""max"",""harness"":""live leaderboard snapshot as of 2026-07-23"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""leaderboard snapshot 2026-07-23"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: AgentArena = 9.8. Source setting: live leaderboard snapshot as of 2026-07-23. Origin: third-party leaderboard cited by Kimi report.
gpt-5.5,critpt,27.1,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""max/xhigh; no tools"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Kimi K3 technical report: CritPt = 27.1. Source setting: max/xhigh; no tools. Origin: Artificial Analysis cited by Kimi report.
gpt-5.5,aa_lcr,74.3,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""max/xhigh; no tools"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Kimi K3 technical report: AA-LCR = 74.3. Source setting: max/xhigh; no tools. Origin: Artificial Analysis cited by Kimi report.
gpt-5.5,hle_tools,52.2,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""general tools; pass@1"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""general tools""}",True,tech_report,verified,Kimi K3 technical report: HLE-Full (with tools) = 52.2. Source setting: general tools; pass@1. Origin: Moonshot evaluation.
gpt-5.5,deep_swe_v1_1,67.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""agentic coding harness per row"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: DeepSWE = 67.0. Source setting: agentic coding harness per row. Origin: Moonshot K3 run; competitors from DeepSWE leaderboard / GLM blog.
gpt-5.5,terminal_bench_2_1,83.4,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""agentic coding harness per row"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Terminal-Bench 2.1 = 83.4. Source setting: agentic coding harness per row. Origin: mixed official sources documented by Kimi report.
gpt-5.5,frontier_swe,64.9,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""agentic coding harness per row"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: FrontierSWE = 64.9. Source setting: agentic coding harness per row. Origin: Moonshot/OpenAI runs plus official leaderboard.
gpt-5.5,swe_marathon_h20_2026_07_09,14.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""H20-calibrated pre-final-v1.1 branch (2026-07-09)"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: SWE-Marathon = 14.0. Source setting: H20-calibrated pre-final-v1.1 branch (2026-07-09). Origin: Moonshot evaluation.
gpt-5.5,posttrain_bench,28.4,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""H20 avg@3 for K3/Fable/Sol; official H100 for Opus/GPT-5.5/GLM"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: PostTrainBench = 28.4. Source setting: H20 avg@3 for K3/Fable/Sol; official H100 for Opus/GPT-5.5/GLM. Origin: Moonshot H20 runs plus official H100 leaderboard values.
gpt-5.5,scicode,56.1,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""agentic coding harness per row"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: SciCode = 56.1. Source setting: agentic coding harness per row. Origin: Artificial Analysis cited by Kimi report.
gpt-5.5,researchrubrics,64.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: ResearchRubrics = 64.0. Source setting: agentic benchmark harness. Origin: Moonshot evaluation.
gpt-5.5,gdpval_aa_elo,1491.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: GDPval-AA v2 (Elo) = 1491.0. Source setting: agentic benchmark harness. Origin: Artificial Analysis cited by Kimi report.
gpt-5.5,toolathlon_verified,73.5,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""Verified edition; official leaderboard snapshot"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Toolathlon-Verified = 73.5. Source setting: Verified edition; official leaderboard snapshot. Origin: official Toolathlon leaderboard cited by Kimi report.
gpt-5.5,automation_bench,22.7,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: AutomationBench = 22.7. Source setting: agentic benchmark harness. Origin: Moonshot evaluation.
gpt-5.5,job_bench,38.3,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: JobBench = 38.3. Source setting: agentic benchmark harness. Origin: official JobBench leaderboard cited by Kimi report.
gpt-5.5,aa_briefcase_elo,1158.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: AA-Briefcase (Elo) = 1158.0. Source setting: agentic benchmark harness. Origin: Artificial Analysis cited by Kimi report.
gpt-5.5,agents_last_exam,26.6,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Agents' Last Exam = 26.6. Source setting: agentic benchmark harness. Origin: official leaderboard cited by Kimi report.
gpt-5.5,apex_agents,38.5,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: APEX-Agents = 38.5. Source setting: agentic benchmark harness. Origin: Artificial Analysis cited by Kimi report.
gpt-5.5,spreadsheetbench_2,29.1,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: SpreadsheetBench 2 = 29.1. Source setting: agentic benchmark harness. Origin: Moonshot evaluation.
gpt-5.5,osworld_verified,79.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: OSWorld-Verified = 79.0. Source setting: agentic benchmark harness. Origin: Moonshot evaluation.
gpt-5.5,osworld_2_0,49.5,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: OSWorld 2.0 = 49.5. Source setting: agentic benchmark harness. Origin: Moonshot evaluation.
gpt-5.5,saas_bench,43.8,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: SaaS-Bench = 43.8. Source setting: agentic benchmark harness. Origin: Moonshot evaluation.
gpt-5.5,tau3_banking,31.3,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: τ³-Banking = 31.3. Source setting: agentic benchmark harness. Origin: Artificial Analysis cited by Kimi report.
gpt-5.5,harvey_lab_aa,86.3,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Harvey Lab-AA = 86.3. Source setting: agentic benchmark harness. Origin: Artificial Analysis cited by Kimi report.
gpt-5.5,corpfin_v2,68.4,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: CorpFin v2 = 68.4. Source setting: agentic benchmark harness. Origin: Vals AI cited by Kimi report.
gpt-5.5,finance_agent_v2,51.8,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Finance Agent v2 = 51.8. Source setting: agentic benchmark harness. Origin: Vals AI cited by Kimi report.
gpt-5.5,legal_research_bench,40.4,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Legal Research Bench = 40.4. Source setting: agentic benchmark harness. Origin: Vals AI cited by Kimi report.
gpt-5.5,worldvqa,38.5,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""forced-answer prompt; avg 3 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,Kimi K3 technical report: WorldVQA ForceAnswer = 38.5. Source setting: forced-answer prompt; avg 3 runs. Origin: Moonshot evaluation.
gpt-5.5,omnidocbench,0.106,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""avg 3 runs; source 1-NED score normalized to NED"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: OmniDocBench = 0.106 NED. Source setting: avg 3 runs; source 1-NED score normalized to NED. Origin: Moonshot evaluation. The source reports 1-NED=89.4%; stored value is normalized NED=0.106.
gpt-5.5,perception_bench,55.8,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""multimodal; avg 3 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: PerceptionBench = 55.8. Source setting: multimodal; avg 3 runs. Origin: Moonshot evaluation.
gpt-5.5,video_mme,89.3,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""audio+visual+subtitles; avg 3 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,Kimi K3 technical report: Video-MME (w. sub) = 89.3. Source setting: audio+visual+subtitles; avg 3 runs. Origin: Moonshot evaluation.
gpt-5.5,mmvu,81.7,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""avg 3 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: MMVU = 81.7. Source setting: avg 3 runs. Origin: Moonshot evaluation.
gpt-5.5,babyvision,83.6,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""Python tool; avg 3 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""python""}",False,tech_report,verified,Kimi K3 technical report: BabyVision w/ python = 83.6. Source setting: Python tool; avg 3 runs. Origin: Moonshot evaluation.
gpt-5.5,charxiv_reasoning,84.1,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""vision; avg 3 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,Kimi K3 technical report: CharXiv (RQ) (no Python) = 84.1. Source setting: vision; avg 3 runs. Origin: Moonshot evaluation.
gpt-5.5,mathvision,92.2,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""vision; avg 3 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""avg-of-3"",""temperature"":""default"",""tools"":""none""}",False,tech_report,verified,Kimi K3 technical report: MathVision (no Python) = 92.2. Source setting: vision; avg 3 runs. Origin: Moonshot evaluation.
gpt-5.5,zerobench_main,22.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""no Python; pass@5 / 5 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@5"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: ZeroBench (pass@5) (no Python) = 22.0. Source setting: no Python; pass@5 / 5 runs. Origin: Moonshot evaluation. Canonical compatibility remains under review; stored as best available primary with matches_canonical=false.
gpt-5.5,zerobench_tools,41.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""Python tool; pass@5 / 5 runs"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@5"",""temperature"":""default"",""tools"":""python""}",False,tech_report,verified,Kimi K3 technical report: ZeroBench (pass@5) (with Python) = 41.0. Source setting: Python tool; pass@5 / 5 runs. Origin: Moonshot evaluation. Canonical compatibility remains under review; stored as best available primary with matches_canonical=false.
gpt-5.5,coding_experience,56.8,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""provider-aligned harness: Kimi Code / Claude Code / Codex"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Coding Experience = 56.8. Source setting: provider-aligned harness: Kimi Code / Claude Code / Codex. Origin: Moonshot evaluation.
gpt-5.5,clawbench_2_0,48.5,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""OpenClaw harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: 24/7 ClawBench 2.0 = 48.5. Source setting: OpenClaw harness. Origin: Moonshot evaluation.
gpt-5.5,mira_bench,54.6,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""MIRA harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: MIRA Bench = 54.6. Source setting: MIRA harness. Origin: Moonshot evaluation.
gpt-5.5,kaet,79.7,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""Kimi Code harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: KAET = 79.7. Source setting: Kimi Code harness. Origin: Moonshot evaluation.
gpt-5.5,clif_bench,52.3,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""Kimi Code harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: CLIF Bench = 52.3. Source setting: Kimi Code harness. Origin: Moonshot evaluation.
gpt-5.5,agentic_vision_bench,76.9,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""Kimi Code harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Agentic Vision Bench = 76.9. Source setting: Kimi Code harness. Origin: Moonshot evaluation.
gpt-5.5,swarm_bench,61.8,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""Kimi Agent harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: SwarmBench = 61.8. Source setting: Kimi Agent harness. Origin: Moonshot evaluation.
gpt-5.5,online_experience,73.7,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""Kimi Agent harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Online Experience = 73.7. Source setting: Kimi Agent harness. Origin: Moonshot evaluation.
gpt-5.5,deepresearchbench,81.9,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""Kimi Agent harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: DeepResearchBench = 81.9. Source setting: Kimi Agent harness. Origin: Moonshot evaluation.
gpt-5.5,finance_bench,58.4,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""N/A harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Finance Bench = 58.4. Source setting: N/A harness. Origin: Moonshot evaluation.
gpt-5.5,kwv_bench,65.8,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""N/A harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: KWVBench = 65.8. Source setting: N/A harness. Origin: Moonshot evaluation.
gpt-5.5,deck_bench,68.2,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""N/A harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: DECKBench = 68.2. Source setting: N/A harness. Origin: Moonshot evaluation.
gpt-5.5,agent_behavior_bench,70.1,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""Kimi Work harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Agent Behavior Bench = 70.1. Source setting: Kimi Work harness. Origin: Moonshot evaluation.
gpt-5.5,faithfulness,86.5,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""N/A harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Faithfulness (1 - hallucination rate) = 86.5. Source setting: N/A harness. Origin: Moonshot evaluation.
gpt-5.5,chat_all_in_one_bench,71.8,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""Kimi Work harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Chat All-in-One Bench = 71.8. Source setting: Kimi Work harness. Origin: Moonshot evaluation.
gpt-5.5,aa_intelligence_index_v4_1,55.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""live leaderboard snapshot as of 2026-07-23"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""leaderboard snapshot 2026-07-23"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: AA Intelligence Index v4.1 = 55.0. Source setting: live leaderboard snapshot as of 2026-07-23. Origin: third-party leaderboard cited by Kimi report.
gpt-5.5,vals_index,68.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""live leaderboard snapshot as of 2026-07-23"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""leaderboard snapshot 2026-07-23"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Vals Index = 68.0. Source setting: live leaderboard snapshot as of 2026-07-23. Origin: third-party leaderboard cited by Kimi report.
gpt-5.5,webdev_arena_elo,1507.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""live leaderboard snapshot as of 2026-07-23"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""leaderboard snapshot 2026-07-23"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: WebDevArena Elo = 1507. Source setting: live leaderboard snapshot as of 2026-07-23. Origin: third-party leaderboard cited by Kimi report.
gpt-5.5,text_arena_elo,1482.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""high"",""harness"":""live leaderboard snapshot as of 2026-07-23"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""leaderboard snapshot 2026-07-23"",""temperature"":""default"",""tools"":""benchmark-specified""}",False,tech_report,verified,Kimi K3 technical report: TextArena Elo = 1482. Source setting: live leaderboard snapshot as of 2026-07-23. Origin: third-party leaderboard cited by Kimi report.
gpt-5.5,agent_arena,8.8,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""default"",""effort"":""xhigh"",""harness"":""live leaderboard snapshot as of 2026-07-23"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""leaderboard snapshot 2026-07-23"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: AgentArena = 8.8. Source setting: live leaderboard snapshot as of 2026-07-23. Origin: third-party leaderboard cited by Kimi report.
glm-5.2,gpqa_diamond,91.2,https://z.ai/blog/glm-5.2,"{""context"":""default"",""effort"":""max"",""harness"":""official reasoning evaluation"",""judge"":""GPT-5.5 medium for AIME/HMMT/IMO; benchmark-specified otherwise"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none"",""top_p"":""0.95""}",True,official_blog,verified,GLM-5.2 official release source: GLM-5.2 / GPQA-Diamond = 91.2.
glm-5.2,critpt,20.9,https://z.ai/blog/glm-5.2,"{""context"":""default"",""effort"":""max"",""harness"":""official reasoning evaluation"",""judge"":""GPT-5.5 medium for AIME/HMMT/IMO; benchmark-specified otherwise"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none"",""top_p"":""0.95""}",True,official_blog,verified,GLM-5.2 official release source: GLM-5.2 / CritPt = 20.9.
glm-5.2,aa_lcr,71.3,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""max/xhigh; no tools"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none""}",True,tech_report,verified,Kimi K3 technical report: AA-LCR = 71.3. Source setting: max/xhigh; no tools. Origin: Artificial Analysis cited by Kimi report.
glm-5.2,deep_swe_v1_1,46.2,https://z.ai/blog/glm-5.2,"{""context"":""400000"",""effort"":""max"",""harness"":""official pier + mini-swe-agent; timeout=2h; 2 CPU/8GB; no internet"",""judge"":""isolated-container verification"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic code execution"",""top_p"":""1.0""}",True,official_blog,verified,GLM-5.2 official release source: GLM-5.2 / DeepSWE = 46.2.
glm-5.2,program_bench,63.7,https://z.ai/blog/glm-5.2,"{""context"":""400000"",""effort"":""max"",""harness"":""Claude Code 2.1.156; max_turns=2000; timeout=6h; 4 CPU/8GB; no internet"",""judge"":""248,000+ fuzz-generated behavioral tests"",""max_output_tokens"":""64000"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic code execution"",""top_p"":""1.0""}",True,official_blog,verified,GLM-5.2 official release source: GLM-5.2 / ProgramBench = 63.7.
glm-5.2,terminal_bench_2_1,82.7,https://z.ai/blog/glm-5.2,"{""context"":""source-reported"",""effort"":""max"",""harness"":""Claude Code 2.1.167; Claude Code values averaged over 5 runs"",""judge"":""official task verifier"",""max_output_tokens"":""131072 for Claude Code; source-reported otherwise"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0 for Claude Code; unknown otherwise"",""tools"":""terminal agent"",""top_p"":""0.95 for Claude Code; unknown otherwise""}",True,official_blog,verified,GLM-5.2 official release source: GLM-5.2 / Terminal Bench 2.1 (Best Reported Harness) = 82.7.
glm-5.2,frontier_swe,74.4,https://z.ai/blog/glm-5.2,"{""context"":""1000000"",""effort"":""max"",""harness"":""Proximal evaluation; max effort; as of 2026-06-16"",""judge"":""FrontierSWE dominance"",""max_output_tokens"":""128000"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic code execution""}",True,official_blog,verified,GLM-5.2 official release source: GLM-5.2 / FrontierSWE (Dominance) = 74.4.
glm-5.2,swe_marathon_h20_2026_07_09,13.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""H20-calibrated pre-final-v1.1 branch (2026-07-09)"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: SWE-Marathon = 13.0. Source setting: H20-calibrated pre-final-v1.1 branch (2026-07-09). Origin: Moonshot evaluation.
glm-5.2,posttrain_bench,34.3,https://z.ai/blog/glm-5.2,"{""context"":""1000000"",""effort"":""max"",""harness"":""PostTrainBench evaluation; max effort"",""judge"":""official weighted benchmark score"",""max_output_tokens"":""128000"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic post-training""}",True,official_blog,verified,GLM-5.2 official release source: GLM-5.2 / PostTrainBench = 34.3.
glm-5.2,mls_bench_lite,40.4,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic coding harness per row"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: MLS-Bench-Lite = 40.4. Source setting: agentic coding harness per row. Origin: Moonshot evaluation.
glm-5.2,scicode,50.5,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic coding harness per row"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: SciCode = 50.5. Source setting: agentic coding harness per row. Origin: Artificial Analysis cited by Kimi report.
glm-5.2,kimi_code_bench_v2,64.2,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic coding harness per row"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Kimi Code Bench 2.0 = 64.2. Source setting: agentic coding harness per row. Origin: Moonshot evaluation.
glm-5.2,researchrubrics,71.1,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: ResearchRubrics = 71.1. Source setting: agentic benchmark harness. Origin: Moonshot evaluation.
glm-5.2,gdpval_aa_elo,1510.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: GDPval-AA v2 (Elo) = 1510.0. Source setting: agentic benchmark harness. Origin: Artificial Analysis cited by Kimi report.
glm-5.2,toolathlon_verified,59.9,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""Verified edition; official leaderboard snapshot"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Toolathlon-Verified = 59.9. Source setting: Verified edition; official leaderboard snapshot. Origin: official Toolathlon leaderboard cited by Kimi report.
glm-5.2,mcpatlas,76.8,https://z.ai/blog/glm-5.2,"{""context"":""default"",""effort"":""max"",""harness"":""public 500-task set; 10-minute timeout per task"",""judge"":""Gemini 3.0 Pro"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic MCP servers""}",False,official_blog,verified,GLM-5.2 official release source: GLM-5.2 / MCP-Atlas (Public Set) = 76.8.
glm-5.2,automation_bench,12.9,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: AutomationBench = 12.9. Source setting: agentic benchmark harness. Origin: Moonshot evaluation.
glm-5.2,job_bench,43.4,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: JobBench = 43.4. Source setting: agentic benchmark harness. Origin: official JobBench leaderboard cited by Kimi report.
glm-5.2,aa_briefcase_elo,1260.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: AA-Briefcase (Elo) = 1260.0. Source setting: agentic benchmark harness. Origin: Artificial Analysis cited by Kimi report.
glm-5.2,agents_last_exam,20.4,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Agents' Last Exam = 20.4. Source setting: agentic benchmark harness. Origin: official leaderboard cited by Kimi report.
glm-5.2,apex_agents,35.6,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: APEX-Agents = 35.6. Source setting: agentic benchmark harness. Origin: Artificial Analysis cited by Kimi report.
glm-5.2,officeqa_pro,41.4,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""PDF corpus rendered as images; no machine-readable text"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: OfficeQA Pro = 41.4. Source setting: PDF corpus rendered as images; no machine-readable text. Origin: Moonshot evaluation.
glm-5.2,spreadsheetbench_2,28.1,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: SpreadsheetBench 2 = 28.1. Source setting: agentic benchmark harness. Origin: Moonshot evaluation.
glm-5.2,tau3_banking,26.8,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: τ³-Banking = 26.8. Source setting: agentic benchmark harness. Origin: Artificial Analysis cited by Kimi report.
glm-5.2,harvey_lab_aa,91.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Harvey Lab-AA = 91.0. Source setting: agentic benchmark harness. Origin: Artificial Analysis cited by Kimi report.
glm-5.2,corpfin_v2,66.1,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: CorpFin v2 = 66.1. Source setting: agentic benchmark harness. Origin: Vals AI cited by Kimi report.
glm-5.2,finance_agent_v2,49.7,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Finance Agent v2 = 49.7. Source setting: agentic benchmark harness. Origin: Vals AI cited by Kimi report.
glm-5.2,legal_research_bench,31.3,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""agentic benchmark harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Legal Research Bench = 31.3. Source setting: agentic benchmark harness. Origin: Vals AI cited by Kimi report.
glm-5.2,coding_experience,53.3,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""provider-aligned harness: Kimi Code / Claude Code / Codex"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Coding Experience = 53.3. Source setting: provider-aligned harness: Kimi Code / Claude Code / Codex. Origin: Moonshot evaluation.
glm-5.2,clawbench_2_0,43.2,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""OpenClaw harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: 24/7 ClawBench 2.0 = 43.2. Source setting: OpenClaw harness. Origin: Moonshot evaluation.
glm-5.2,kaet,74.7,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""Kimi Code harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: KAET = 74.7. Source setting: Kimi Code harness. Origin: Moonshot evaluation.
glm-5.2,clif_bench,39.2,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""Kimi Code harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: CLIF Bench = 39.2. Source setting: Kimi Code harness. Origin: Moonshot evaluation.
glm-5.2,swarm_bench,58.5,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""Kimi Agent harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: SwarmBench = 58.5. Source setting: Kimi Agent harness. Origin: Moonshot evaluation.
glm-5.2,online_experience,64.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""Kimi Agent harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Online Experience = 64.0. Source setting: Kimi Agent harness. Origin: Moonshot evaluation.
glm-5.2,deepresearchbench,84.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""Kimi Agent harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: DeepResearchBench = 84.0. Source setting: Kimi Agent harness. Origin: Moonshot evaluation.
glm-5.2,finance_bench,55.4,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""N/A harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Finance Bench = 55.4. Source setting: N/A harness. Origin: Moonshot evaluation.
glm-5.2,deck_bench,68.6,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""N/A harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: DECKBench = 68.6. Source setting: N/A harness. Origin: Moonshot evaluation.
glm-5.2,faithfulness,74.8,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""N/A harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Faithfulness (1 - hallucination rate) = 74.8. Source setting: N/A harness. Origin: Moonshot evaluation.
glm-5.2,aa_intelligence_index_v4_1,51.1,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""live leaderboard snapshot as of 2026-07-23"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""leaderboard snapshot 2026-07-23"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: AA Intelligence Index v4.1 = 51.1. Source setting: live leaderboard snapshot as of 2026-07-23. Origin: third-party leaderboard cited by Kimi report.
glm-5.2,vals_index,65.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""live leaderboard snapshot as of 2026-07-23"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""leaderboard snapshot 2026-07-23"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Vals Index = 65.0. Source setting: live leaderboard snapshot as of 2026-07-23. Origin: third-party leaderboard cited by Kimi report.
glm-5.2,webdev_arena_elo,1592.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""live leaderboard snapshot as of 2026-07-23"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""leaderboard snapshot 2026-07-23"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: WebDevArena Elo = 1592. Source setting: live leaderboard snapshot as of 2026-07-23. Origin: third-party leaderboard cited by Kimi report.
glm-5.2,text_arena_elo,1469.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""live leaderboard snapshot as of 2026-07-23"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""leaderboard snapshot 2026-07-23"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: TextArena Elo = 1469. Source setting: live leaderboard snapshot as of 2026-07-23. Origin: third-party leaderboard cited by Kimi report.
glm-5.2,agent_arena,6.5,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""live leaderboard snapshot as of 2026-07-23"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""leaderboard snapshot 2026-07-23"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: AgentArena = 6.5. Source setting: live leaderboard snapshot as of 2026-07-23. Origin: third-party leaderboard cited by Kimi report.
glm-5.2,moonshot_exploit_development_suite,22.2,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""36-task in-house exploit suite"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: Moonshot exploit-development suite = 22.2. Source setting: 36-task in-house exploit suite. Origin: Moonshot evaluation.
glm-5.2,exploitbench,24.0,https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""independent joint assessment"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""benchmark-specified""}",True,tech_report,verified,Kimi K3 technical report: ExploitBench (UK AISI / NIST CAISI) = 24.0. Source setting: independent joint assessment. Origin: UK AISI / NIST CAISI cited by Kimi report.
deepseek-v4-flash-0731,terminal_bench_2_1,82.7,https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""DeepSeek Harness minimal mode for Code Agent tasks"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95"",""tools"":""benchmark-specified""}",True,model_card,verified,DeepSeek-V4-Flash-0731 official model card: DeepSeek-V4-Flash-0731 / Terminal Bench 2.1 = 82.7. Cross-model settings are marked unknown when the source does not restate them.
deepseek-v4-flash,terminal_bench_2_1,61.8,https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731,"{""context"":""unknown"",""effort"":""unknown"",""harness"":""not restated in DeepSeek-V4-Flash-0731 model card"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""benchmark-specified""}",False,model_card,verified,DeepSeek-V4-Flash-0731 official model card: DeepSeek-V4-Flash (Preview) / Terminal Bench 2.1 = 61.8. Cross-model settings are marked unknown when the source does not restate them.
deepseek-v4-pro,terminal_bench_2_1,72.1,https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731,"{""context"":""unknown"",""effort"":""unknown"",""harness"":""not restated in DeepSeek-V4-Flash-0731 model card"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""benchmark-specified""}",False,model_card,verified,DeepSeek-V4-Flash-0731 official model card: DeepSeek-V4-Pro (Preview) / Terminal Bench 2.1 = 72.1. Cross-model settings are marked unknown when the source does not restate them.
deepseek-v4-flash-0731,nl2repo_bench,54.2,https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""DeepSeek Harness minimal mode for Code Agent tasks"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95"",""tools"":""benchmark-specified""}",True,model_card,verified,DeepSeek-V4-Flash-0731 official model card: DeepSeek-V4-Flash-0731 / NL2Repo = 54.2. Cross-model settings are marked unknown when the source does not restate them.
deepseek-v4-flash,nl2repo_bench,39.4,https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731,"{""context"":""unknown"",""effort"":""unknown"",""harness"":""not restated in DeepSeek-V4-Flash-0731 model card"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""benchmark-specified""}",False,model_card,verified,DeepSeek-V4-Flash-0731 official model card: DeepSeek-V4-Flash (Preview) / NL2Repo = 39.4. Cross-model settings are marked unknown when the source does not restate them.
deepseek-v4-pro,nl2repo_bench,38.5,https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731,"{""context"":""unknown"",""effort"":""unknown"",""harness"":""not restated in DeepSeek-V4-Flash-0731 model card"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""benchmark-specified""}",False,model_card,verified,DeepSeek-V4-Flash-0731 official model card: DeepSeek-V4-Pro (Preview) / NL2Repo = 38.5. Cross-model settings are marked unknown when the source does not restate them.
glm-5.2,nl2repo_bench,48.9,https://z.ai/blog/glm-5.2,"{""context"":""400000"",""effort"":""max"",""harness"":""official NL2Repo evaluation"",""judge"":""rule-based + LLM anti-hack judgement"",""max_output_tokens"":""48000"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic code execution"",""top_p"":""1.0""}",True,official_blog,verified,GLM-5.2 official release source: GLM-5.2 / NL2Repo = 48.9.
claude-opus-4.8,nl2repo_bench,69.7,https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731,"{""context"":""unknown"",""effort"":""unknown"",""harness"":""not restated in DeepSeek-V4-Flash-0731 model card"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""benchmark-specified""}",False,model_card,verified,DeepSeek-V4-Flash-0731 official model card: Opus-4.8 / NL2Repo = 69.7. Cross-model settings are marked unknown when the source does not restate them.
deepseek-v4-flash-0731,cybergym,76.7,https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""DeepSeek Harness minimal mode for Code Agent tasks"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95"",""tools"":""benchmark-specified""}",True,model_card,verified,DeepSeek-V4-Flash-0731 official model card: DeepSeek-V4-Flash-0731 / Cybergym = 76.7. Cross-model settings are marked unknown when the source does not restate them.
deepseek-v4-flash,cybergym,38.7,https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731,"{""context"":""unknown"",""effort"":""unknown"",""harness"":""not restated in DeepSeek-V4-Flash-0731 model card"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""benchmark-specified""}",False,model_card,verified,DeepSeek-V4-Flash-0731 official model card: DeepSeek-V4-Flash (Preview) / Cybergym = 38.7. Cross-model settings are marked unknown when the source does not restate them.
deepseek-v4-pro,cybergym,52.7,https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731,"{""context"":""unknown"",""effort"":""unknown"",""harness"":""not restated in DeepSeek-V4-Flash-0731 model card"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""benchmark-specified""}",False,model_card,verified,DeepSeek-V4-Flash-0731 official model card: DeepSeek-V4-Pro (Preview) / Cybergym = 52.7. Cross-model settings are marked unknown when the source does not restate them.
claude-opus-4.8,cybergym,83.1,https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731,"{""context"":""unknown"",""effort"":""unknown"",""harness"":""not restated in DeepSeek-V4-Flash-0731 model card"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""benchmark-specified""}",False,model_card,verified,DeepSeek-V4-Flash-0731 official model card: Opus-4.8 / Cybergym = 83.1. Cross-model settings are marked unknown when the source does not restate them.
deepseek-v4-flash-0731,deep_swe_v1_1,54.4,https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731,"{""context"":""1048576 (1M)"",""effort"":""max"",""harness"":""DeepSeek Harness minimal mode for Code Agent tasks"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=0.95"",""tools"":""benchmark-specified""}",True,model_card,verified,DeepSeek-V4-Flash-0731 official model card: DeepSeek-V4-Flash-0731 / DeepSWE = 54.4. Cross-model settings are marked unknown when the source does not restate them.
deepseek-v4-flash,deep_swe_v1_1,7.3,https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731,"{""context"":""unknown"",""effort"":""unknown"",""harness"":""not restated in DeepSeek-V4-Flash-0731 model card"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""benchmark-specified""}",False,model_card,verified,DeepSeek-V4-Flash-0731 official model card: DeepSeek-V4-Flash (Preview) / DeepSWE = 7.3. Cross-model settings are marked unknown when the source does not restate them.
deepseek-v4-pro,deep_swe_v1_1,12.8,https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731,"{""context"":""unknown"",""effort"":""unknown"",""harness"":""not restated in DeepSeek-V4-Flash-0731 model card"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""benchmark-specified""}",False,model_card,verified,DeepSeek-V4-Flash-0731 official model card: DeepSeek-V4-Pro (Preview) / DeepSWE = 12.8. Cross-model settings are marked unknown when the source does not restate them.
deepseek-v4-flash-0731,toolathlon_verified,70.3,https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731,"{""context"":""unknown"",""effort"":""unknown"",""harness"":""not restated in DeepSeek-V4-Flash-0731 model card"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""benchmark-specified""}",False,model_card,verified,DeepSeek-V4-Flash-0731 official model card: DeepSeek-V4-Flash-0731 / Toolathlon-Verified = 70.3. Cross-model settings are marked unknown when the source does not restate them.
deepseek-v4-flash,toolathlon_verified,49.7,https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731,"{""context"":""unknown"",""effort"":""unknown"",""harness"":""not restated in DeepSeek-V4-Flash-0731 model card"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""benchmark-specified""}",False,model_card,verified,DeepSeek-V4-Flash-0731 official model card: DeepSeek-V4-Flash (Preview) / Toolathlon-Verified = 49.7. Cross-model settings are marked unknown when the source does not restate them.
deepseek-v4-pro,toolathlon_verified,55.9,https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731,"{""context"":""unknown"",""effort"":""unknown"",""harness"":""not restated in DeepSeek-V4-Flash-0731 model card"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""benchmark-specified""}",False,model_card,verified,DeepSeek-V4-Flash-0731 official model card: DeepSeek-V4-Pro (Preview) / Toolathlon-Verified = 55.9. Cross-model settings are marked unknown when the source does not restate them.
deepseek-v4-flash-0731,agents_last_exam,25.2,https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731,"{""context"":""unknown"",""effort"":""unknown"",""harness"":""not restated in DeepSeek-V4-Flash-0731 model card"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""benchmark-specified""}",False,model_card,verified,DeepSeek-V4-Flash-0731 official model card: DeepSeek-V4-Flash-0731 / Agents' Last Exam = 25.2. Cross-model settings are marked unknown when the source does not restate them.
deepseek-v4-flash,agents_last_exam,15.8,https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731,"{""context"":""unknown"",""effort"":""unknown"",""harness"":""not restated in DeepSeek-V4-Flash-0731 model card"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""benchmark-specified""}",False,model_card,verified,DeepSeek-V4-Flash-0731 official model card: DeepSeek-V4-Flash (Preview) / Agents' Last Exam = 15.8. Cross-model settings are marked unknown when the source does not restate them.
deepseek-v4-pro,agents_last_exam,16.5,https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731,"{""context"":""unknown"",""effort"":""unknown"",""harness"":""not restated in DeepSeek-V4-Flash-0731 model card"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""benchmark-specified""}",False,model_card,verified,DeepSeek-V4-Flash-0731 official model card: DeepSeek-V4-Pro (Preview) / Agents' Last Exam = 16.5. Cross-model settings are marked unknown when the source does not restate them.
deepseek-v4-flash-0731,automation_bench,25.1,https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731,"{""context"":""unknown"",""effort"":""unknown"",""harness"":""not restated in DeepSeek-V4-Flash-0731 model card"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""benchmark-specified""}",False,model_card,verified,DeepSeek-V4-Flash-0731 official model card: DeepSeek-V4-Flash-0731 / AutomationBench Public = 25.1. Cross-model settings are marked unknown when the source does not restate them.
deepseek-v4-flash,automation_bench,10.8,https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731,"{""context"":""unknown"",""effort"":""unknown"",""harness"":""not restated in DeepSeek-V4-Flash-0731 model card"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""benchmark-specified""}",False,model_card,verified,DeepSeek-V4-Flash-0731 official model card: DeepSeek-V4-Flash (Preview) / AutomationBench Public = 10.8. Cross-model settings are marked unknown when the source does not restate them.
deepseek-v4-pro,automation_bench,12.8,https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731,"{""context"":""unknown"",""effort"":""unknown"",""harness"":""not restated in DeepSeek-V4-Flash-0731 model card"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""benchmark-specified""}",False,model_card,verified,DeepSeek-V4-Flash-0731 official model card: DeepSeek-V4-Pro (Preview) / AutomationBench Public = 12.8. Cross-model settings are marked unknown when the source does not restate them.
deepseek-v4-flash-0731,dsbench_fullstack,68.7,https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731,"{""context"":""unknown"",""effort"":""unknown"",""harness"":""not restated in DeepSeek-V4-Flash-0731 model card"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""benchmark-specified""}",False,model_card,verified,DeepSeek-V4-Flash-0731 official model card: DeepSeek-V4-Flash-0731 / DSBench-FullStack † = 68.7. Cross-model settings are marked unknown when the source does not restate them.
deepseek-v4-flash,dsbench_fullstack,37.0,https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731,"{""context"":""unknown"",""effort"":""unknown"",""harness"":""not restated in DeepSeek-V4-Flash-0731 model card"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""benchmark-specified""}",False,model_card,verified,DeepSeek-V4-Flash-0731 official model card: DeepSeek-V4-Flash (Preview) / DSBench-FullStack † = 37.0. Cross-model settings are marked unknown when the source does not restate them.
deepseek-v4-pro,dsbench_fullstack,41.8,https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731,"{""context"":""unknown"",""effort"":""unknown"",""harness"":""not restated in DeepSeek-V4-Flash-0731 model card"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""benchmark-specified""}",False,model_card,verified,DeepSeek-V4-Flash-0731 official model card: DeepSeek-V4-Pro (Preview) / DSBench-FullStack † = 41.8. Cross-model settings are marked unknown when the source does not restate them.
glm-5.2,dsbench_fullstack,61.8,https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731,"{""context"":""unknown"",""effort"":""unknown"",""harness"":""not restated in DeepSeek-V4-Flash-0731 model card"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""benchmark-specified""}",False,model_card,verified,DeepSeek-V4-Flash-0731 official model card: GLM-5.2 / DSBench-FullStack † = 61.8. Cross-model settings are marked unknown when the source does not restate them.
claude-opus-4.8,dsbench_fullstack,71.6,https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731,"{""context"":""unknown"",""effort"":""unknown"",""harness"":""not restated in DeepSeek-V4-Flash-0731 model card"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""benchmark-specified""}",False,model_card,verified,DeepSeek-V4-Flash-0731 official model card: Opus-4.8 / DSBench-FullStack † = 71.6. Cross-model settings are marked unknown when the source does not restate them.
deepseek-v4-flash-0731,dsbench_hard,59.6,https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731,"{""context"":""unknown"",""effort"":""unknown"",""harness"":""not restated in DeepSeek-V4-Flash-0731 model card"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""benchmark-specified""}",False,model_card,verified,DeepSeek-V4-Flash-0731 official model card: DeepSeek-V4-Flash-0731 / DSBench-Hard † = 59.6. Cross-model settings are marked unknown when the source does not restate them.
deepseek-v4-flash,dsbench_hard,25.8,https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731,"{""context"":""unknown"",""effort"":""unknown"",""harness"":""not restated in DeepSeek-V4-Flash-0731 model card"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""benchmark-specified""}",False,model_card,verified,DeepSeek-V4-Flash-0731 official model card: DeepSeek-V4-Flash (Preview) / DSBench-Hard † = 25.8. Cross-model settings are marked unknown when the source does not restate them.
deepseek-v4-pro,dsbench_hard,31.1,https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731,"{""context"":""unknown"",""effort"":""unknown"",""harness"":""not restated in DeepSeek-V4-Flash-0731 model card"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""benchmark-specified""}",False,model_card,verified,DeepSeek-V4-Flash-0731 official model card: DeepSeek-V4-Pro (Preview) / DSBench-Hard † = 31.1. Cross-model settings are marked unknown when the source does not restate them.
glm-5.2,dsbench_hard,54.5,https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731,"{""context"":""unknown"",""effort"":""unknown"",""harness"":""not restated in DeepSeek-V4-Flash-0731 model card"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""benchmark-specified""}",False,model_card,verified,DeepSeek-V4-Flash-0731 official model card: GLM-5.2 / DSBench-Hard † = 54.5. Cross-model settings are marked unknown when the source does not restate them.
claude-opus-4.8,dsbench_hard,71.7,https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731,"{""context"":""unknown"",""effort"":""unknown"",""harness"":""not restated in DeepSeek-V4-Flash-0731 model card"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""benchmark-specified""}",False,model_card,verified,DeepSeek-V4-Flash-0731 official model card: Opus-4.8 / DSBench-Hard † = 71.7. Cross-model settings are marked unknown when the source does not restate them.
glm-5.2,hle_text,40.5,https://z.ai/blog/glm-5.2,"{""context"":""default"",""effort"":""max"",""harness"":""official reasoning evaluation"",""judge"":""GPT-5.5 medium for AIME/HMMT/IMO; benchmark-specified otherwise"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none"",""top_p"":""0.95""}",True,official_blog,verified,GLM-5.2 official release source: GLM-5.2 / HLE = 40.5.
glm-5.2,hle_tools_text,54.7,https://z.ai/blog/glm-5.2,"{""context"":""300000"",""effort"":""max"",""harness"":""official; no context-management strategy"",""judge"":""answer-key scoring"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""search+code+web"",""top_p"":""0.95""}",True,official_blog,verified,GLM-5.2 official release source: GLM-5.2 / HLE (w/ Tools) = 54.7.
glm-5.2,aime_2026,99.2,https://z.ai/blog/glm-5.2,"{""context"":""default"",""effort"":""max"",""harness"":""official reasoning evaluation"",""judge"":""GPT-5.5 medium for AIME/HMMT/IMO; benchmark-specified otherwise"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none"",""top_p"":""0.95""}",True,official_blog,verified,GLM-5.2 official release source: GLM-5.2 / AIME 2026 = 99.2.
glm-5.2,hmmt_nov_2025,94.4,https://z.ai/blog/glm-5.2,"{""context"":""default"",""effort"":""max"",""harness"":""official reasoning evaluation"",""judge"":""GPT-5.5 medium for AIME/HMMT/IMO; benchmark-specified otherwise"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none"",""top_p"":""0.95""}",True,official_blog,verified,GLM-5.2 official release source: GLM-5.2 / HMMT Nov. 2025 = 94.4.
glm-5.2,hmmt_feb_2026,92.5,https://z.ai/blog/glm-5.2,"{""context"":""default"",""effort"":""max"",""harness"":""official reasoning evaluation"",""judge"":""GPT-5.5 medium for AIME/HMMT/IMO; benchmark-specified otherwise"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none"",""top_p"":""0.95""}",True,official_blog,verified,GLM-5.2 official release source: GLM-5.2 / HMMT Feb. 2026 = 92.5.
glm-5.2,imo_answerbench,91.0,https://z.ai/blog/glm-5.2,"{""context"":""default"",""effort"":""max"",""harness"":""official reasoning evaluation"",""judge"":""GPT-5.5 medium for AIME/HMMT/IMO; benchmark-specified otherwise"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none"",""top_p"":""0.95""}",True,official_blog,verified,GLM-5.2 official release source: GLM-5.2 / IMOAnswerBench = 91.0.
glm-5.2,swe_bench_pro,62.1,https://z.ai/blog/glm-5.2,"{""context"":""400000"",""effort"":""max"",""harness"":""OpenHands with tailored instruction prompt"",""judge"":""official tests"",""max_output_tokens"":""32000"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic code execution"",""top_p"":""1.0""}",True,official_blog,verified,GLM-5.2 official release source: GLM-5.2 / SWE-bench Pro = 62.1.
glm-5.2,swe_marathon_v1_0,13.0,https://z.ai/blog/glm-5.2,"{""context"":""1000000"",""effort"":""max"",""harness"":""Abundant AI SWE-Marathon v1.0; max effort"",""judge"":""official task verifiers"",""max_output_tokens"":""128000"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic code execution""}",True,official_blog,verified,GLM-5.2 official release source: GLM-5.2 / SWE-Marathon = 13.0.
glm-5.2,toolathlon,48.2,https://z.ai/blog/glm-5.2,"{""context"":""default"",""effort"":""max"",""harness"":""original Toolathlon public evaluation service (pre-Verified)"",""judge"":""official evaluation service"",""max_output_tokens"":""128000"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic tool use""}",True,official_blog,verified,GLM-5.2 official release source: GLM-5.2 / Tool-Decathlon = 48.2.
glm-5.1,hle_tools_text,52.3,https://z.ai/blog/glm-5.2,"{""context"":""300000"",""effort"":""official reported"",""harness"":""official; no context-management strategy"",""judge"":""answer-key scoring"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""search+code+web"",""top_p"":""0.95""}",True,official_blog,verified,GLM-5.2 official release source: GLM-5.1 / HLE (w/ Tools) = 52.3.
glm-5.1,critpt,4.6,https://z.ai/blog/glm-5.2,"{""context"":""default"",""effort"":""official reported"",""harness"":""official reasoning evaluation"",""judge"":""GPT-5.5 medium for AIME/HMMT/IMO; benchmark-specified otherwise"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none"",""top_p"":""0.95""}",True,official_blog,verified,GLM-5.2 official release source: GLM-5.1 / CritPt = 4.6.
glm-5.1,deep_swe_v1_1,18.0,https://z.ai/blog/glm-5.2,"{""context"":""400000"",""effort"":""official reported"",""harness"":""official pier + mini-swe-agent; timeout=2h; 2 CPU/8GB; no internet"",""judge"":""isolated-container verification"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic code execution"",""top_p"":""1.0""}",True,official_blog,verified,GLM-5.2 official release source: GLM-5.1 / DeepSWE = 18.0.
glm-5.1,program_bench,50.9,https://z.ai/blog/glm-5.2,"{""context"":""400000"",""effort"":""official reported"",""harness"":""Claude Code 2.1.156; max_turns=2000; timeout=6h; 4 CPU/8GB; no internet"",""judge"":""248,000+ fuzz-generated behavioral tests"",""max_output_tokens"":""64000"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic code execution"",""top_p"":""1.0""}",True,official_blog,verified,GLM-5.2 official release source: GLM-5.1 / ProgramBench = 50.9.
glm-5.1,terminal_bench_2_1,69.0,https://z.ai/blog/glm-5.2,"{""context"":""source-reported"",""effort"":""official reported"",""harness"":""Claude Code 2.1.167; Claude Code values averaged over 5 runs"",""judge"":""official task verifier"",""max_output_tokens"":""131072 for Claude Code; source-reported otherwise"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0 for Claude Code; unknown otherwise"",""tools"":""terminal agent"",""top_p"":""0.95 for Claude Code; unknown otherwise""}",True,official_blog,verified,GLM-5.2 official release source: GLM-5.1 / Terminal Bench 2.1 (Best Reported Harness) = 69.0.
glm-5.1,frontier_swe,30.5,https://z.ai/blog/glm-5.2,"{""context"":""1000000"",""effort"":""official reported"",""harness"":""Proximal evaluation; max effort; as of 2026-06-16"",""judge"":""FrontierSWE dominance"",""max_output_tokens"":""128000"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic code execution""}",True,official_blog,verified,GLM-5.2 official release source: GLM-5.1 / FrontierSWE (Dominance) = 30.5.
glm-5.1,posttrain_bench,20.1,https://z.ai/blog/glm-5.2,"{""context"":""1000000"",""effort"":""official reported"",""harness"":""PostTrainBench evaluation; max effort"",""judge"":""official weighted benchmark score"",""max_output_tokens"":""128000"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic post-training""}",True,official_blog,verified,GLM-5.2 official release source: GLM-5.1 / PostTrainBench = 20.1.
glm-5.1,swe_marathon_v1_0,1.0,https://z.ai/blog/glm-5.2,"{""context"":""1000000"",""effort"":""official reported"",""harness"":""Abundant AI SWE-Marathon v1.0; max effort"",""judge"":""official task verifiers"",""max_output_tokens"":""128000"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic code execution""}",True,official_blog,verified,GLM-5.2 official release source: GLM-5.1 / SWE-Marathon = 1.0.
qwen3.7-max,hle_text,41.4,https://www.alibabacloud.com/blog/qwen3-7-the-agent-frontier_603154,"{""context"":""default"",""effort"":""xhigh / maximum thinking"",""harness"":""official reasoning evaluation"",""judge"":""GPT-5.5 medium for AIME/HMMT/IMO; benchmark-specified otherwise"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none"",""top_p"":""0.95""}",True,official_blog,verified,Qwen3.7-Max first-party source: HLE = 41.4; GLM-5.2 table used as cross-check.
qwen3.7-max,hle_tools_text,53.5,https://z.ai/blog/glm-5.2,"{""context"":""300000"",""effort"":""xhigh / maximum thinking"",""harness"":""official; no context-management strategy"",""judge"":""answer-key scoring"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""search+code+web"",""top_p"":""0.95""}",False,official_blog,verified,GLM-5.2 official release source: Qwen3.7-Max / HLE (w/ Tools) = 53.5.
qwen3.7-max,critpt,13.4,https://z.ai/blog/glm-5.2,"{""context"":""default"",""effort"":""xhigh / maximum thinking"",""harness"":""official reasoning evaluation"",""judge"":""GPT-5.5 medium for AIME/HMMT/IMO; benchmark-specified otherwise"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none"",""top_p"":""0.95""}",False,official_blog,verified,GLM-5.2 official release source: Qwen3.7-Max / CritPt = 13.4.
qwen3.7-max,aime_2026,97.0,https://z.ai/blog/glm-5.2,"{""context"":""default"",""effort"":""xhigh / maximum thinking"",""harness"":""official reasoning evaluation"",""judge"":""GPT-5.5 medium for AIME/HMMT/IMO; benchmark-specified otherwise"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none"",""top_p"":""0.95""}",False,official_blog,verified,GLM-5.2 official release source: Qwen3.7-Max / AIME 2026 = 97.0.
qwen3.7-max,hmmt_nov_2025,95.0,https://z.ai/blog/glm-5.2,"{""context"":""default"",""effort"":""xhigh / maximum thinking"",""harness"":""official reasoning evaluation"",""judge"":""GPT-5.5 medium for AIME/HMMT/IMO; benchmark-specified otherwise"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none"",""top_p"":""0.95""}",False,official_blog,verified,GLM-5.2 official release source: Qwen3.7-Max / HMMT Nov. 2025 = 95.0.
qwen3.7-max,hmmt_feb_2026,97.1,https://www.alibabacloud.com/blog/qwen3-7-the-agent-frontier_603154,"{""context"":""default"",""effort"":""xhigh / maximum thinking"",""harness"":""official reasoning evaluation"",""judge"":""GPT-5.5 medium for AIME/HMMT/IMO; benchmark-specified otherwise"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none"",""top_p"":""0.95""}",True,official_blog,verified,Qwen3.7-Max first-party source: HMMT Feb. 2026 = 97.1; GLM-5.2 table used as cross-check.
qwen3.7-max,imo_answerbench,90.0,https://www.alibabacloud.com/blog/qwen3-7-the-agent-frontier_603154,"{""context"":""default"",""effort"":""xhigh / maximum thinking"",""harness"":""official reasoning evaluation"",""judge"":""GPT-5.5 medium for AIME/HMMT/IMO; benchmark-specified otherwise"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none"",""top_p"":""0.95""}",True,official_blog,verified,Qwen3.7-Max first-party source: IMOAnswerBench = 90.0; GLM-5.2 table used as cross-check.
qwen3.7-max,gpqa_diamond,92.4,https://www.alibabacloud.com/blog/qwen3-7-the-agent-frontier_603154,"{""context"":""default"",""effort"":""xhigh / maximum thinking"",""harness"":""official reasoning evaluation"",""judge"":""GPT-5.5 medium for AIME/HMMT/IMO; benchmark-specified otherwise"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none"",""top_p"":""0.95""}",True,official_blog,verified,Qwen3.7-Max first-party source: GPQA-Diamond = 92.4; GLM-5.2 table used as cross-check.
qwen3.7-max,swe_bench_pro,60.6,https://www.alibabacloud.com/blog/qwen3-7-the-agent-frontier_603154,"{""context"":""400000"",""effort"":""xhigh / maximum thinking"",""harness"":""OpenHands with tailored instruction prompt"",""judge"":""official tests"",""max_output_tokens"":""32000"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic code execution"",""top_p"":""1.0""}",True,official_blog,verified,Qwen3.7-Max first-party source: SWE-bench Pro = 60.6; GLM-5.2 table used as cross-check.
qwen3.7-max,nl2repo_bench,47.2,https://z.ai/blog/glm-5.2,"{""context"":""400000"",""effort"":""xhigh / maximum thinking"",""harness"":""official NL2Repo evaluation"",""judge"":""rule-based + LLM anti-hack judgement"",""max_output_tokens"":""48000"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic code execution"",""top_p"":""1.0""}",False,official_blog,verified,GLM-5.2 official release source: Qwen3.7-Max / NL2Repo = 47.2.
qwen3.7-max,deep_swe_v1_1,18.0,https://z.ai/blog/glm-5.2,"{""context"":""400000"",""effort"":""xhigh / maximum thinking"",""harness"":""official pier + mini-swe-agent; timeout=2h; 2 CPU/8GB; no internet"",""judge"":""isolated-container verification"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic code execution"",""top_p"":""1.0""}",False,official_blog,verified,GLM-5.2 official release source: Qwen3.7-Max / DeepSWE = 18.0.
qwen3.7-max,terminal_bench_2_1,75.0,https://z.ai/blog/glm-5.2,"{""context"":""256000"",""effort"":""xhigh / maximum thinking"",""harness"":""Terminus-2; parser=json; timeout=4h; max_episodes=500; 4 CPU/8GB"",""judge"":""official task verifier"",""max_output_tokens"":""48000"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""terminal agent"",""top_p"":""1.0""}",True,official_blog,verified,GLM-5.2 official release source: Qwen3.7-Max / Terminal Bench 2.1 (Terminus-2) = 75.0.
qwen3.7-max,mcpatlas,76.4,https://www.alibabacloud.com/blog/qwen3-7-the-agent-frontier_603154,"{""context"":""default"",""effort"":""xhigh / maximum thinking"",""harness"":""public 500-task set; 10-minute timeout per task"",""judge"":""Gemini 3.0 Pro"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic MCP servers""}",False,official_blog,verified,Qwen3.7-Max first-party source: MCP-Atlas (Public Set) = 76.4; GLM-5.2 table used as cross-check.
minimax-m3,hle_text,37.0,https://z.ai/blog/glm-5.2,"{""context"":""default"",""effort"":""thinking enabled/adaptive"",""harness"":""official reasoning evaluation"",""judge"":""GPT-5.5 medium for AIME/HMMT/IMO; benchmark-specified otherwise"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none"",""top_p"":""0.95""}",False,official_blog,verified,GLM-5.2 official release source: MiniMax M3 / HLE = 37.0.
minimax-m3,critpt,3.7,https://z.ai/blog/glm-5.2,"{""context"":""default"",""effort"":""thinking enabled/adaptive"",""harness"":""official reasoning evaluation"",""judge"":""GPT-5.5 medium for AIME/HMMT/IMO; benchmark-specified otherwise"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none"",""top_p"":""0.95""}",False,official_blog,verified,GLM-5.2 official release source: MiniMax M3 / CritPt = 3.7.
minimax-m3,hmmt_nov_2025,84.4,https://z.ai/blog/glm-5.2,"{""context"":""default"",""effort"":""thinking enabled/adaptive"",""harness"":""official reasoning evaluation"",""judge"":""GPT-5.5 medium for AIME/HMMT/IMO; benchmark-specified otherwise"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none"",""top_p"":""0.95""}",False,official_blog,verified,GLM-5.2 official release source: MiniMax M3 / HMMT Nov. 2025 = 84.4.
minimax-m3,hmmt_feb_2026,84.4,https://z.ai/blog/glm-5.2,"{""context"":""default"",""effort"":""thinking enabled/adaptive"",""harness"":""official reasoning evaluation"",""judge"":""GPT-5.5 medium for AIME/HMMT/IMO; benchmark-specified otherwise"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none"",""top_p"":""0.95""}",False,official_blog,verified,GLM-5.2 official release source: MiniMax M3 / HMMT Feb. 2026 = 84.4.
minimax-m3,gpqa_diamond,93.0,https://z.ai/blog/glm-5.2,"{""context"":""default"",""effort"":""thinking enabled/adaptive"",""harness"":""official reasoning evaluation"",""judge"":""GPT-5.5 medium for AIME/HMMT/IMO; benchmark-specified otherwise"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none"",""top_p"":""0.95""}",False,official_blog,verified,GLM-5.2 official release source: MiniMax M3 / GPQA-Diamond = 93.0.
minimax-m3,swe_bench_pro,59.0,https://www.minimax.io/blog/minimax-m3,"{""context"":""400000"",""effort"":""thinking enabled/adaptive"",""harness"":""OpenHands with tailored instruction prompt"",""judge"":""official tests"",""max_output_tokens"":""32000"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic code execution"",""top_p"":""1.0""}",True,official_blog,verified,MiniMax M3 first-party source: SWE-bench Pro = 59.0; GLM-5.2 table used as cross-check.
minimax-m3,nl2repo_bench,42.1,https://z.ai/blog/glm-5.2,"{""context"":""400000"",""effort"":""thinking enabled/adaptive"",""harness"":""official NL2Repo evaluation"",""judge"":""rule-based + LLM anti-hack judgement"",""max_output_tokens"":""48000"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic code execution"",""top_p"":""1.0""}",False,official_blog,verified,GLM-5.2 official release source: MiniMax M3 / NL2Repo = 42.1.
minimax-m3,deep_swe_v1_1,20.0,https://z.ai/blog/glm-5.2,"{""context"":""400000"",""effort"":""thinking enabled/adaptive"",""harness"":""official pier + mini-swe-agent; timeout=2h; 2 CPU/8GB; no internet"",""judge"":""isolated-container verification"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic code execution"",""top_p"":""1.0""}",False,official_blog,verified,GLM-5.2 official release source: MiniMax M3 / DeepSWE = 20.0.
minimax-m3,terminal_bench_2_1,66.0,https://www.minimax.io/blog/minimax-m3,"{""context"":""256000"",""effort"":""thinking enabled/adaptive"",""harness"":""Terminus-2; parser=json; timeout=4h; max_episodes=500; 4 CPU/8GB"",""judge"":""official task verifier"",""max_output_tokens"":""48000"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""terminal agent"",""top_p"":""1.0""}",True,official_blog,verified,MiniMax M3 first-party source: Terminal Bench 2.1 (Terminus-2) = 66.0; GLM-5.2 table used as cross-check.
minimax-m3,mcpatlas,74.2,https://www.minimax.io/blog/minimax-m3,"{""context"":""default"",""effort"":""thinking enabled/adaptive"",""harness"":""public 500-task set; 10-minute timeout per task"",""judge"":""Gemini 3.0 Pro"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic MCP servers""}",False,official_blog,verified,MiniMax M3 first-party source: MCP-Atlas (Public Set) = 74.2; GLM-5.2 table used as cross-check.
deepseek-v4-pro,hle_tools_text,48.2,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro,"{""context"":""300000"",""effort"":""Max"",""harness"":""official; no context-management strategy"",""judge"":""answer-key scoring"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""search+code+web"",""top_p"":""0.95""}",True,model_card,verified,DeepSeek-V4-Pro first-party source: HLE (w/ Tools) = 48.2; GLM-5.2 table used as cross-check.
deepseek-v4-pro,critpt,12.9,https://z.ai/blog/glm-5.2,"{""context"":""default"",""effort"":""Max"",""harness"":""official reasoning evaluation"",""judge"":""GPT-5.5 medium for AIME/HMMT/IMO; benchmark-specified otherwise"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none"",""top_p"":""0.95""}",True,official_blog,verified,GLM-5.2 official release source: DeepSeek-V4-Pro / CritPt = 12.9.
deepseek-v4-pro,aime_2026,94.6,https://z.ai/blog/glm-5.2,"{""context"":""default"",""effort"":""Max"",""harness"":""official reasoning evaluation"",""judge"":""GPT-5.5 medium for AIME/HMMT/IMO; benchmark-specified otherwise"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none"",""top_p"":""0.95""}",True,official_blog,verified,GLM-5.2 official release source: DeepSeek-V4-Pro / AIME 2026 = 94.6.
deepseek-v4-pro,hmmt_nov_2025,94.4,https://z.ai/blog/glm-5.2,"{""context"":""default"",""effort"":""Max"",""harness"":""official reasoning evaluation"",""judge"":""GPT-5.5 medium for AIME/HMMT/IMO; benchmark-specified otherwise"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none"",""top_p"":""0.95""}",True,official_blog,verified,GLM-5.2 official release source: DeepSeek-V4-Pro / HMMT Nov. 2025 = 94.4.
deepseek-v4-pro,program_bench,47.8,https://z.ai/blog/glm-5.2,"{""context"":""400000"",""effort"":""Max"",""harness"":""Claude Code 2.1.156; max_turns=2000; timeout=6h; 4 CPU/8GB; no internet"",""judge"":""248,000+ fuzz-generated behavioral tests"",""max_output_tokens"":""64000"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic code execution"",""top_p"":""1.0""}",False,official_blog,verified,GLM-5.2 official release source: DeepSeek-V4-Pro / ProgramBench = 47.8.
deepseek-v4-pro,frontier_swe,29.0,https://z.ai/blog/glm-5.2,"{""context"":""1000000"",""effort"":""Max"",""harness"":""Proximal evaluation; max effort; as of 2026-06-16"",""judge"":""FrontierSWE dominance"",""max_output_tokens"":""128000"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic code execution""}",False,official_blog,verified,GLM-5.2 official release source: DeepSeek-V4-Pro / FrontierSWE (Dominance) = 29.0.
claude-opus-4.8,aime_2026,95.7,https://z.ai/blog/glm-5.2,"{""context"":""default"",""effort"":""max (source; model default is high)"",""harness"":""official reasoning evaluation"",""judge"":""GPT-5.5 medium for AIME/HMMT/IMO; benchmark-specified otherwise"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none"",""top_p"":""0.95""}",False,official_blog,verified,GLM-5.2 official release source: Claude Opus 4.8 / AIME 2026 = 95.7.
claude-opus-4.8,hmmt_nov_2025,96.5,https://z.ai/blog/glm-5.2,"{""context"":""default"",""effort"":""max (source; model default is high)"",""harness"":""official reasoning evaluation"",""judge"":""GPT-5.5 medium for AIME/HMMT/IMO; benchmark-specified otherwise"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none"",""top_p"":""0.95""}",False,official_blog,verified,GLM-5.2 official release source: Claude Opus 4.8 / HMMT Nov. 2025 = 96.5.
claude-opus-4.8,hmmt_feb_2026,96.7,https://z.ai/blog/glm-5.2,"{""context"":""default"",""effort"":""max (source; model default is high)"",""harness"":""official reasoning evaluation"",""judge"":""GPT-5.5 medium for AIME/HMMT/IMO; benchmark-specified otherwise"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none"",""top_p"":""0.95""}",False,official_blog,verified,GLM-5.2 official release source: Claude Opus 4.8 / HMMT Feb. 2026 = 96.7.
claude-opus-4.8,imo_answerbench,83.5,https://z.ai/blog/glm-5.2,"{""context"":""default"",""effort"":""max (source; model default is high)"",""harness"":""official reasoning evaluation"",""judge"":""GPT-5.5 medium for AIME/HMMT/IMO; benchmark-specified otherwise"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none"",""top_p"":""0.95""}",False,official_blog,verified,GLM-5.2 official release source: Claude Opus 4.8 / IMOAnswerBench = 83.5.
claude-opus-4.8,swe_bench_pro,69.2,https://z.ai/blog/glm-5.2,"{""context"":""400000"",""effort"":""max (source; model default is high)"",""harness"":""OpenHands with tailored instruction prompt"",""judge"":""official tests"",""max_output_tokens"":""32000"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic code execution"",""top_p"":""1.0""}",False,official_blog,verified,GLM-5.2 official release source: Claude Opus 4.8 / SWE-bench Pro = 69.2.
claude-opus-4.8,swe_marathon_v1_0,26.0,https://z.ai/blog/glm-5.2,"{""context"":""1000000"",""effort"":""max (source; model default is high)"",""harness"":""Abundant AI SWE-Marathon v1.0; max effort"",""judge"":""official task verifiers"",""max_output_tokens"":""128000"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic code execution""}",False,official_blog,verified,GLM-5.2 official release source: Claude Opus 4.8 / SWE-Marathon = 26.0.
claude-opus-4.8,toolathlon,59.9,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""toolathlon"",""sampling"":""3 trials/task"",""tools"":""agentic benchmark harness""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 8.11.5.A, page 135: Claude Opus 4.8; Toolathlon [Pass@1]=59.9. Source setting: effort=max; tools=agentic benchmark harness; sampling=3 trials/task; harness=toolathlon."
gpt-5.5,aime_2026,98.3,https://z.ai/blog/glm-5.2,"{""context"":""default"",""effort"":""xhigh"",""harness"":""official reasoning evaluation"",""judge"":""GPT-5.5 medium for AIME/HMMT/IMO; benchmark-specified otherwise"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none"",""top_p"":""0.95""}",True,official_blog,verified,GLM-5.2 official release source: GPT-5.5 / AIME 2026 = 98.3.
gpt-5.5,hmmt_nov_2025,96.5,https://z.ai/blog/glm-5.2,"{""context"":""default"",""effort"":""xhigh"",""harness"":""official reasoning evaluation"",""judge"":""GPT-5.5 medium for AIME/HMMT/IMO; benchmark-specified otherwise"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none"",""top_p"":""0.95""}",True,official_blog,verified,GLM-5.2 official release source: GPT-5.5 / HMMT Nov. 2025 = 96.5.
gpt-5.5,hmmt_feb_2026,96.7,https://z.ai/blog/glm-5.2,"{""context"":""default"",""effort"":""xhigh"",""harness"":""official reasoning evaluation"",""judge"":""GPT-5.5 medium for AIME/HMMT/IMO; benchmark-specified otherwise"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""none"",""top_p"":""0.95""}",True,official_blog,verified,GLM-5.2 official release source: GPT-5.5 / HMMT Feb. 2026 = 96.7.
gpt-5.5,nl2repo_bench,50.7,https://z.ai/blog/glm-5.2,"{""context"":""400000"",""effort"":""xhigh"",""harness"":""official NL2Repo evaluation"",""judge"":""rule-based + LLM anti-hack judgement"",""max_output_tokens"":""48000"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic code execution"",""top_p"":""1.0""}",True,official_blog,verified,GLM-5.2 official release source: GPT-5.5 / NL2Repo = 50.7.
gpt-5.5,swe_marathon_v1_0,12.0,https://z.ai/blog/glm-5.2,"{""context"":""1000000"",""effort"":""xhigh"",""harness"":""Abundant AI SWE-Marathon v1.0; max effort"",""judge"":""official task verifiers"",""max_output_tokens"":""128000"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic code execution""}",True,official_blog,verified,GLM-5.2 official release source: GPT-5.5 / SWE-Marathon = 12.0.
gemini-3.1-pro,hle_tools,51.4,https://z.ai/blog/glm-5.2,"{""context"":""300000"",""effort"":""high"",""harness"":""official; no context-management strategy"",""judge"":""answer-key scoring"",""max_output_tokens"":""163840"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""search+code+web"",""top_p"":""0.95""}",True,official_blog,verified,GLM-5.2 official release source: Gemini 3.1 Pro / HLE (w/ Tools) = 51.4.
gemini-3.1-pro,deep_swe_v1_1,10.0,https://z.ai/blog/glm-5.2,"{""context"":""400000"",""effort"":""high"",""harness"":""official pier + mini-swe-agent; timeout=2h; 2 CPU/8GB; no internet"",""judge"":""isolated-container verification"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic code execution"",""top_p"":""1.0""}",True,official_blog,verified,GLM-5.2 official release source: Gemini 3.1 Pro / DeepSWE = 10.0.
gemini-3.1-pro,program_bench,39.5,https://z.ai/blog/glm-5.2,"{""context"":""400000"",""effort"":""high"",""harness"":""Claude Code 2.1.156; max_turns=2000; timeout=6h; 4 CPU/8GB; no internet"",""judge"":""248,000+ fuzz-generated behavioral tests"",""max_output_tokens"":""64000"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""agentic code execution"",""top_p"":""1.0""}",True,official_blog,verified,GLM-5.2 official release source: Gemini 3.1 Pro / ProgramBench = 39.5.
gemini-3.1-pro,terminal_bench_2_1,74.0,https://z.ai/blog/glm-5.2,"{""context"":""256000"",""effort"":""high"",""harness"":""Terminus-2; parser=json; timeout=4h; max_episodes=500; 4 CPU/8GB"",""judge"":""official task verifier"",""max_output_tokens"":""48000"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0"",""tools"":""terminal agent"",""top_p"":""1.0""}",True,official_blog,verified,GLM-5.2 official release source: Gemini 3.1 Pro / Terminal Bench 2.1 (Terminus-2) = 74.0.
gemini-3.1-pro,frontier_swe,39.6,https://z.ai/blog/glm-5.2,"{""context"":""1000000"",""effort"":""high"",""harness"":""Proximal evaluation; max effort; as of 2026-06-16"",""judge"":""FrontierSWE dominance"",""max_output_tokens"":""128000"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic code execution""}",True,official_blog,verified,GLM-5.2 official release source: Gemini 3.1 Pro / FrontierSWE (Dominance) = 39.6.
gemini-3.1-pro,posttrain_bench,21.6,https://z.ai/blog/glm-5.2,"{""context"":""1000000"",""effort"":""high"",""harness"":""PostTrainBench evaluation; max effort"",""judge"":""official weighted benchmark score"",""max_output_tokens"":""128000"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic post-training""}",True,official_blog,verified,GLM-5.2 official release source: Gemini 3.1 Pro / PostTrainBench = 21.6.
gemini-3.1-pro,swe_marathon_v1_0,4.0,https://z.ai/blog/glm-5.2,"{""context"":""1000000"",""effort"":""high"",""harness"":""Abundant AI SWE-Marathon v1.0; max effort"",""judge"":""official task verifiers"",""max_output_tokens"":""128000"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic code execution""}",True,official_blog,verified,GLM-5.2 official release source: Gemini 3.1 Pro / SWE-Marathon = 4.0.
claude-opus-4.7,frontier_swe,63.0,https://z.ai/blog/glm-5.2,"{""context"":""1000000"",""effort"":""max (source; model default differs)"",""harness"":""Proximal evaluation; max effort; as of 2026-06-16"",""judge"":""FrontierSWE dominance"",""max_output_tokens"":""128000"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic code execution""}",False,official_blog,verified,GLM-5.2 official release source: Claude Opus 4.7 / FrontierSWE (Dominance) = 63.0.
claude-opus-4.7,posttrain_bench,28.6,https://z.ai/blog/glm-5.2,"{""context"":""1000000"",""effort"":""max (source; model default differs)"",""harness"":""PostTrainBench evaluation; max effort"",""judge"":""official weighted benchmark score"",""max_output_tokens"":""128000"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic post-training""}",False,official_blog,verified,GLM-5.2 official release source: Claude Opus 4.7 / PostTrainBench = 28.6.
claude-opus-4.7,swe_marathon_v1_0,16.0,https://z.ai/blog/glm-5.2,"{""context"":""1000000"",""effort"":""max (source; model default differs)"",""harness"":""Abundant AI SWE-Marathon v1.0; max effort"",""judge"":""official task verifiers"",""max_output_tokens"":""128000"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""agentic code execution""}",False,official_blog,verified,GLM-5.2 official release source: Claude Opus 4.7 / SWE-Marathon = 16.0.
hy3-preview,swe_bench_multilingual,68.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""SWE-agent scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",True,model_card,verified,Hy3 official model-card appendix: SWE-bench Multilingual = 68.3. Comparator-reported value.
hy3,swe_bench_multilingual,75.8,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""SWE-agent scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",True,model_card,verified,Hy3 official model-card appendix: SWE-bench Multilingual = 75.8. Comparator-reported value.
glm-5.2,swe_bench_multilingual,83.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""SWE-agent scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",False,model_card,verified,Hy3 official model-card appendix: SWE-bench Multilingual = 83.0*. Tencent own testing.
doubao-seed-2.1-pro,swe_bench_multilingual,75.1,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""SWE-agent scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",False,model_card,verified,Hy3 official model-card appendix: SWE-bench Multilingual = 75.1*. Tencent own testing.
qwen3.7-max,swe_bench_multilingual,78.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""SWE-agent scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",False,model_card,verified,Hy3 official model-card appendix: SWE-bench Multilingual = 78.3. Comparator-reported value.
claude-opus-4.8,swe_bench_multilingual,84.8,https://longcat.chat/blog/longcat-2.0/,"{""context"":""default"",""effort"":""high"",""harness"":""provider official report; details not reproduced by LongCat"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Asterisk: LongCat cites this value from the model provider's official report rather than measuring it in-house."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""provider official setting"",""tools"":""agentic code execution""}",True,official_blog,verified,LongCat-2.0 official tech blog: SWE-bench Multilingual = 84.8*. Asterisked value is cited by LongCat from the model provider's official report.
gpt-5.5,swe_bench_multilingual,77.8,https://media.x.ai/v1/website/4p5-5184fdf9.pdf,"{""context"":""unknown"",""effort"":""xhigh"",""harness"":""fixed SWE-bench Multilingual agent scaffold"",""judge"":""language-appropriate isolated tests"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic repository shell/editor""}",True,model_card,verified,card45 SWE-Bench Multilingual Resolve rate (%); source effort=xhigh; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
hy3-preview,swe_bench_verified,74.4,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""SWE-agent scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",True,model_card,verified,Hy3 official model-card appendix: SWE-bench Verified = 74.4. Comparator-reported value.
hy3,swe_bench_verified,78.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""SWE-agent scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",True,model_card,verified,Hy3 official model-card appendix: SWE-bench Verified = 78.0. Comparator-reported value.
glm-5.1,swe_bench_verified,75.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""SWE-agent scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",False,model_card,verified,Hy3 official model-card appendix: SWE-bench Verified = 75.0*. Tencent own testing.
glm-5.2,swe_bench_verified,84.2,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""SWE-agent scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",False,model_card,verified,Hy3 official model-card appendix: SWE-bench Verified = 84.2*. Tencent own testing.
doubao-seed-2.1-pro,swe_bench_verified,76.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""SWE-agent scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",False,model_card,verified,Hy3 official model-card appendix: SWE-bench Verified = 76.0*. Tencent own testing.
qwen3.7-max,swe_bench_verified,80.4,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""SWE-agent scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",False,model_card,verified,Hy3 official model-card appendix: SWE-bench Verified = 80.4. Comparator-reported value.
claude-opus-4.8,swe_bench_verified,88.6,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""SWE-agent scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",False,model_card,verified,Hy3 official model-card appendix: SWE-bench Verified = 88.6. Comparator-reported value.
gpt-5.5,swe_bench_verified,84.4,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""SWE-agent scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",False,model_card,verified,Hy3 official model-card appendix: SWE-bench Verified = 84.4*. Tencent own testing.
hy3-preview,swe_bench_pro,46.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""SWE-agent scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",True,model_card,verified,Hy3 official model-card appendix: SWE-bench Pro = 46.0. Comparator-reported value.
hy3,swe_bench_pro,57.9,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""SWE-agent scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",True,model_card,verified,Hy3 official model-card appendix: SWE-bench Pro = 57.9. Comparator-reported value.
doubao-seed-2.1-pro,swe_bench_pro,57.5,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",True,official_blog,verified,Release blog EN: Coding chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4wlbt.png).
hy3-preview,terminal_bench_2_1,58.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Terminus-2; parser=xml; timeout=4h; CPU=16 cores; memory=32GB; max episodes=500"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""terminal/code execution""}",True,model_card,verified,Hy3 official model-card appendix: Terminal-Bench 2.1 = 58.0. Comparator-reported value.
hy3,terminal_bench_2_1,71.7,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Terminus-2; parser=xml; timeout=4h; CPU=16 cores; memory=32GB; max episodes=500"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""terminal/code execution""}",True,model_card,verified,Hy3 official model-card appendix: Terminal-Bench 2.1 = 71.7. Comparator-reported value.
doubao-seed-2.1-pro,terminal_bench_2_1,71.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",True,official_blog,verified,Release blog EN: Coding chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4wlbt.png).
hy3-preview,nl2repo_bench,35.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Claude Code; 250-turn budget; timeout=12000s; 4 CPU/32GB; anti-hacking/tool-call monitoring"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""Claude Code agentic code execution""}",True,model_card,verified,Hy3 official model-card appendix: NL2Repo = 35.3. Comparator-reported value.
hy3,nl2repo_bench,45.6,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Claude Code; 250-turn budget; timeout=12000s; 4 CPU/32GB; anti-hacking/tool-call monitoring"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""Claude Code agentic code execution""}",True,model_card,verified,Hy3 official model-card appendix: NL2Repo = 45.6. Comparator-reported value.
doubao-seed-2.1-pro,nl2repo_bench,47.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",True,official_blog,verified,Release blog EN: Coding chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4wlbt.png).
hy3-preview,deep_swe_v1_1,0.9,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""mini-swe-agent; 2h/task; 2 CPU/8GB; no internet"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",True,model_card,verified,Hy3 official model-card appendix: DeepSWE = 0.9. Comparator-reported value.
hy3,deep_swe_v1_1,28.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""mini-swe-agent; 2h/task; 2 CPU/8GB; no internet"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",True,model_card,verified,Hy3 official model-card appendix: DeepSWE = 28.0. Comparator-reported value.
doubao-seed-2.1-pro,deep_swe_v1_1,32.7,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",True,official_blog,verified,Release blog EN: Coding chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4wlbt.png).
hy3-preview,hy_backend_2_0,12.2,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Claude Code scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",True,model_card,verified,Hy3 official model-card appendix: Hy-Backend 2.0 (internal) = 12.2. Comparator-reported value.
hy3,hy_backend_2_0,25.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Claude Code scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",True,model_card,verified,Hy3 official model-card appendix: Hy-Backend 2.0 (internal) = 25.0. Comparator-reported value.
glm-5.1,hy_backend_2_0,24.7,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Claude Code scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",False,model_card,verified,Hy3 official model-card appendix: Hy-Backend 2.0 (internal) = 24.7*. Tencent own testing.
glm-5.2,hy_backend_2_0,30.4,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Claude Code scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",False,model_card,verified,Hy3 official model-card appendix: Hy-Backend 2.0 (internal) = 30.4*. Tencent own testing.
deepseek-v4-flash,hy_backend_2_0,16.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Claude Code scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",False,model_card,verified,Hy3 official model-card appendix: Hy-Backend 2.0 (internal) = 16.0*. Tencent own testing.
deepseek-v4-pro,hy_backend_2_0,17.5,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Claude Code scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",False,model_card,verified,Hy3 official model-card appendix: Hy-Backend 2.0 (internal) = 17.5*. Tencent own testing.
doubao-seed-2.1-pro,hy_backend_2_0,23.7,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Claude Code scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",False,model_card,verified,Hy3 official model-card appendix: Hy-Backend 2.0 (internal) = 23.7*. Tencent own testing.
qwen3.7-max,hy_backend_2_0,24.7,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Claude Code scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",False,model_card,verified,Hy3 official model-card appendix: Hy-Backend 2.0 (internal) = 24.7*. Tencent own testing.
claude-opus-4.8,hy_backend_2_0,38.9,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Claude Code scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",False,model_card,verified,Hy3 official model-card appendix: Hy-Backend 2.0 (internal) = 38.9*. Tencent own testing.
gpt-5.5,hy_backend_2_0,38.7,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""CodeX scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",False,model_card,verified,Hy3 official model-card appendix: Hy-Backend 2.0 (internal) = 38.7*. Tencent own testing.
hy3-preview,hy_swe_max,30.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Claude Code scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",True,model_card,verified,Hy3 official model-card appendix: Hy-SWE Max (internal) = 30.0. Comparator-reported value.
hy3,hy_swe_max,49.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Claude Code scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",True,model_card,verified,Hy3 official model-card appendix: Hy-SWE Max (internal) = 49.0. Comparator-reported value.
glm-5.1,hy_swe_max,46.1,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Claude Code scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",False,model_card,verified,Hy3 official model-card appendix: Hy-SWE Max (internal) = 46.1*. Tencent own testing.
glm-5.2,hy_swe_max,63.1,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Claude Code scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",False,model_card,verified,Hy3 official model-card appendix: Hy-SWE Max (internal) = 63.1*. Tencent own testing.
deepseek-v4-flash,hy_swe_max,40.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Claude Code scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",False,model_card,verified,Hy3 official model-card appendix: Hy-SWE Max (internal) = 40.0*. Tencent own testing.
deepseek-v4-pro,hy_swe_max,43.1,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Claude Code scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",False,model_card,verified,Hy3 official model-card appendix: Hy-SWE Max (internal) = 43.1*. Tencent own testing.
doubao-seed-2.1-pro,hy_swe_max,55.1,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Claude Code scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",False,model_card,verified,Hy3 official model-card appendix: Hy-SWE Max (internal) = 55.1*. Tencent own testing.
qwen3.7-max,hy_swe_max,48.8,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Claude Code scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",False,model_card,verified,Hy3 official model-card appendix: Hy-SWE Max (internal) = 48.8*. Tencent own testing.
claude-opus-4.8,hy_swe_max,63.2,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Claude Code scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",False,model_card,verified,Hy3 official model-card appendix: Hy-SWE Max (internal) = 63.2*. Tencent own testing.
gpt-5.5,hy_swe_max,65.4,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""CodeX scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",False,model_card,verified,Hy3 official model-card appendix: Hy-SWE Max (internal) = 65.4*. Tencent own testing.
hy3-preview,hy_company_bench,8.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Claude Code scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",True,model_card,verified,Hy3 official model-card appendix: Hy-CompanyBench (internal) = 8.3. Comparator-reported value.
hy3,hy_company_bench,41.7,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Claude Code scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",True,model_card,verified,Hy3 official model-card appendix: Hy-CompanyBench (internal) = 41.7. Comparator-reported value.
glm-5.1,hy_company_bench,28.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Claude Code scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",False,model_card,verified,Hy3 official model-card appendix: Hy-CompanyBench (internal) = 28.3*. Tencent own testing.
glm-5.2,hy_company_bench,55.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Claude Code scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",False,model_card,verified,Hy3 official model-card appendix: Hy-CompanyBench (internal) = 55.0*. Tencent own testing.
deepseek-v4-flash,hy_company_bench,28.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Claude Code scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",False,model_card,verified,Hy3 official model-card appendix: Hy-CompanyBench (internal) = 28.3*. Tencent own testing.
deepseek-v4-pro,hy_company_bench,18.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Claude Code scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",False,model_card,verified,Hy3 official model-card appendix: Hy-CompanyBench (internal) = 18.3*. Tencent own testing.
doubao-seed-2.1-pro,hy_company_bench,20.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Claude Code scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",False,model_card,verified,Hy3 official model-card appendix: Hy-CompanyBench (internal) = 20.0*. Tencent own testing.
qwen3.7-max,hy_company_bench,43.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Claude Code scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",False,model_card,verified,Hy3 official model-card appendix: Hy-CompanyBench (internal) = 43.3*. Tencent own testing.
claude-opus-4.8,hy_company_bench,65.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Claude Code scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",False,model_card,verified,Hy3 official model-card appendix: Hy-CompanyBench (internal) = 65.0*. Tencent own testing.
gpt-5.5,hy_company_bench,51.7,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""CodeX scaffold"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic code execution""}",False,model_card,verified,Hy3 official model-card appendix: Hy-CompanyBench (internal) = 51.7*. Tencent own testing.
hy3-preview,browsecomp,67.1,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal search harness; BrowseComp uses self-summary context management"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""web search/browser""}",True,model_card,verified,Hy3 official model-card appendix: BrowseComp = 67.1. Comparator-reported value.
hy3,browsecomp,84.2,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal search harness; BrowseComp uses self-summary context management"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""web search/browser""}",True,model_card,verified,Hy3 official model-card appendix: BrowseComp = 84.2. Comparator-reported value.
doubao-seed-2.1-pro,browsecomp,86.2,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""search""}",True,official_blog,verified,Release blog EN: Language/search chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq510uc.png).
hy3-preview,widesearch,67.8,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal search harness; BrowseComp uses self-summary context management"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""web search/browser""}",True,model_card,verified,Hy3 official model-card appendix: WideSearch = 67.8. Comparator-reported value.
hy3,widesearch,76.4,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal search harness; BrowseComp uses self-summary context management"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""web search/browser""}",True,model_card,verified,Hy3 official model-card appendix: WideSearch = 76.4. Comparator-reported value.
glm-5.1,widesearch,76.1,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal search harness; BrowseComp uses self-summary context management"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""web search/browser""}",False,model_card,verified,Hy3 official model-card appendix: WideSearch = 76.1*. Tencent own testing.
deepseek-v4-flash,widesearch,74.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal search harness; BrowseComp uses self-summary context management"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""web search/browser""}",False,model_card,verified,Hy3 official model-card appendix: WideSearch = 74.3*. Tencent own testing.
deepseek-v4-pro,widesearch,75.7,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal search harness; BrowseComp uses self-summary context management"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""web search/browser""}",False,model_card,verified,Hy3 official model-card appendix: WideSearch = 75.7*. Tencent own testing.
doubao-seed-2.1-pro,widesearch,76.4,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal search harness; BrowseComp uses self-summary context management"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""web search/browser""}",False,model_card,verified,Hy3 official model-card appendix: WideSearch = 76.4*. Tencent own testing.
qwen3.7-max,widesearch,72.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal search harness; BrowseComp uses self-summary context management"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""web search/browser""}",False,model_card,verified,Hy3 official model-card appendix: WideSearch = 72.3*. Tencent own testing.
gemini-3.1-pro,widesearch,71.7,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal search harness; BrowseComp uses self-summary context management"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""web search/browser""}",False,model_card,verified,Hy3 official model-card appendix: WideSearch = 71.7*. Tencent own testing.
claude-opus-4.8,widesearch,72.9,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal search harness; BrowseComp uses self-summary context management"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""web search/browser""}",False,model_card,verified,Hy3 official model-card appendix: WideSearch = 72.9*. Tencent own testing.
gpt-5.5,widesearch,80.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal search harness; BrowseComp uses self-summary context management"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""web search/browser""}",False,model_card,verified,Hy3 official model-card appendix: WideSearch = 80.0*. Tencent own testing.
hy3-preview,deepsearchqa_f1,82.8,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal search harness; BrowseComp uses self-summary context management"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""web search/browser""}",True,model_card,verified,Hy3 official model-card appendix: DeepSearchQA = 82.8. Comparator-reported value.
hy3,deepsearchqa_f1,91.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal search harness; BrowseComp uses self-summary context management"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""web search/browser""}",True,model_card,verified,Hy3 official model-card appendix: DeepSearchQA = 91.0. Comparator-reported value.
glm-5.1,deepsearchqa_f1,89.1,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal search harness; BrowseComp uses self-summary context management"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""web search/browser""}",False,model_card,verified,Hy3 official model-card appendix: DeepSearchQA = 89.1*. Tencent own testing.
deepseek-v4-flash,deepsearchqa_f1,87.7,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal search harness; BrowseComp uses self-summary context management"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""web search/browser""}",False,model_card,verified,Hy3 official model-card appendix: DeepSearchQA = 87.7*. Tencent own testing.
deepseek-v4-pro,deepsearchqa_f1,90.4,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal search harness; BrowseComp uses self-summary context management"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""web search/browser""}",False,model_card,verified,Hy3 official model-card appendix: DeepSearchQA = 90.4*. Tencent own testing.
doubao-seed-2.1-pro,deepsearchqa_f1,90.8,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal search harness; BrowseComp uses self-summary context management"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""web search/browser""}",False,model_card,verified,Hy3 official model-card appendix: DeepSearchQA = 90.8*. Tencent own testing.
qwen3.7-max,deepsearchqa_f1,88.4,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal search harness; BrowseComp uses self-summary context management"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""web search/browser""}",False,model_card,verified,Hy3 official model-card appendix: DeepSearchQA = 88.4*. Tencent own testing.
gpt-5.5,deepsearchqa_f1,95.5,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal search harness; BrowseComp uses self-summary context management"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""web search/browser""}",False,model_card,verified,Hy3 official model-card appendix: DeepSearchQA = 95.5*. Tencent own testing.
hy3-preview,mcpatlas,66.1,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Scale April 2026 methodology; public 500; 100 tool-call budget"",""judge"":""Gemini 2.5 Pro"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""MCP tool servers""}",True,model_card,verified,Hy3 official model-card appendix: MCP atlas (public) = 66.1. Comparator-reported value.
hy3,mcpatlas,79.1,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Scale April 2026 methodology; public 500; 100 tool-call budget"",""judge"":""Gemini 2.5 Pro"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""MCP tool servers""}",True,model_card,verified,Hy3 official model-card appendix: MCP atlas (public) = 79.1. Comparator-reported value.
doubao-seed-2.1-pro,mcpatlas,83.8,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""MCP servers in an agent harness""}",True,official_blog,verified,Release blog EN: Daily-life/tool-use chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4ujsp.png).
hy3-preview,toolathlon,32.1,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""original Toolathlon official evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic tool use""}",True,model_card,verified,Hy3 official model-card appendix: Toolathlon = 32.1. Comparator-reported value.
hy3,toolathlon,48.5,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""original Toolathlon official evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic tool use""}",True,model_card,verified,Hy3 official model-card appendix: Toolathlon = 48.5. Comparator-reported value.
doubao-seed-2.1-pro,toolathlon,50.6,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic tool use""}",True,official_blog,verified,Release blog EN: Daily-life/tool-use chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4ujsp.png).
qwen3.7-max,toolathlon,45.1,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""original Toolathlon official evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic tool use""}",False,model_card,verified,Hy3 official model-card appendix: Toolathlon = 45.1*. Tencent own testing.
hy3-preview,apex_agents,12.4,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""APEX-Agents long-horizon professional task harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""multiple application/software tools""}",True,model_card,verified,Hy3 official model-card appendix: Apex-Agent (pass@1) = 12.4. Comparator-reported value.
hy3,apex_agents,25.6,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""APEX-Agents long-horizon professional task harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""multiple application/software tools""}",True,model_card,verified,Hy3 official model-card appendix: Apex-Agent (pass@1) = 25.6. Comparator-reported value.
glm-5.1,apex_agents,24.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""APEX-Agents long-horizon professional task harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""multiple application/software tools""}",False,model_card,verified,Hy3 official model-card appendix: Apex-Agent (pass@1) = 24.3*. Tencent own testing.
deepseek-v4-flash,apex_agents,18.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""APEX-Agents long-horizon professional task harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""multiple application/software tools""}",False,model_card,verified,Hy3 official model-card appendix: Apex-Agent (pass@1) = 18.3*. Tencent own testing.
deepseek-v4-pro,apex_agents,20.4,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""APEX-Agents long-horizon professional task harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""multiple application/software tools""}",False,model_card,verified,Hy3 official model-card appendix: Apex-Agent (pass@1) = 20.4*. Tencent own testing.
doubao-seed-2.1-pro,apex_agents,33.8,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""long-horizon professional-task environment""}",True,official_blog,verified,Release blog EN: General-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4u89c.png).
qwen3.7-max,apex_agents,22.2,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""APEX-Agents long-horizon professional task harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""multiple application/software tools""}",False,model_card,verified,Hy3 official model-card appendix: Apex-Agent (pass@1) = 22.2*. Tencent own testing.
hy3-preview,claw_eval_pass3,55.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal harness; 20260325 version; 105 queries"",""judge"":""Gemini 3.5 Flash"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass^3"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic/OpenClaw-style""}",True,model_card,verified,Hy3 official model-card appendix: ClawEval (pass^3) = 55.0. Comparator-reported value.
hy3,claw_eval_pass3,68.5,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal harness; 20260325 version; 105 queries"",""judge"":""Gemini 3.5 Flash"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass^3"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic/OpenClaw-style""}",True,model_card,verified,Hy3 official model-card appendix: ClawEval (pass^3) = 68.5. Comparator-reported value.
glm-5.1,claw_eval_pass3,62.3,https://raw.githubusercontent.com/claw-eval/claw-eval.github.io/e10516b5c2a6dfc139be35dacb54cdf5e034801d/assets/benchmark-BgnHRA8t.js,"{""aggregation_pass_k"":""Pass^3: all three trials must pass"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""Claw-Eval v1.1 non-multimodal: 161 general + 38 multi-turn tasks"",""effort"":""source does not state"",""harness_agent"":""official Claw-Eval v1.1 non-multimodal leaderboard"",""judge"":""full-trajectory completion/safety/robustness grading"",""mode"":""source does not state"",""multimodal_input"":false,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""official Claw-Eval agent environment"",""top_p"":""source does not state"",""trials_samples"":3}",True,official_repository,verified,Official StepFun launch comparison table. Direct locked official score authority reports Claweval-v1.1=62.3; the StepFun surface is retained as physical corroboration.
glm-5.2,claw_eval_pass3,62.4,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal harness; 20260325 version; 105 queries"",""judge"":""Gemini 3.5 Flash"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass^3"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic/OpenClaw-style""}",False,model_card,verified,Hy3 official model-card appendix: ClawEval (pass^3) = 62.4*. Tencent own testing.
deepseek-v4-flash,claw_eval_pass3,61.8,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal harness; 20260325 version; 105 queries"",""judge"":""Gemini 3.5 Flash"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass^3"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic/OpenClaw-style""}",False,model_card,verified,Hy3 official model-card appendix: ClawEval (pass^3) = 61.8*. Tencent own testing.
deepseek-v4-pro,claw_eval_pass3,58.4,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal harness; 20260325 version; 105 queries"",""judge"":""Gemini 3.5 Flash"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass^3"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic/OpenClaw-style""}",False,model_card,verified,Hy3 official model-card appendix: ClawEval (pass^3) = 58.4. Comparator-reported value.
doubao-seed-2.1-pro,claw_eval_pass3,62.1,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal harness; 20260325 version; 105 queries"",""judge"":""Gemini 3.5 Flash"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass^3"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic/OpenClaw-style""}",False,model_card,verified,Hy3 official model-card appendix: ClawEval (pass^3) = 62.1*. Tencent own testing.
qwen3.7-max,claw_eval_pass3,65.2,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal harness; 20260325 version; 105 queries"",""judge"":""Gemini 3.5 Flash"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass^3"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic/OpenClaw-style""}",False,model_card,verified,Hy3 official model-card appendix: ClawEval (pass^3) = 65.2. Comparator-reported value.
gemini-3.1-pro,claw_eval_pass3,50.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal harness; 20260325 version; 105 queries"",""judge"":""Gemini 3.5 Flash"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass^3"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic/OpenClaw-style""}",False,model_card,verified,Hy3 official model-card appendix: ClawEval (pass^3) = 50.0*. Tencent own testing.
claude-opus-4.8,claw_eval_pass3,72.1,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal harness; 20260325 version; 105 queries"",""judge"":""Gemini 3.5 Flash"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass^3"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic/OpenClaw-style""}",False,model_card,verified,Hy3 official model-card appendix: ClawEval (pass^3) = 72.1*. Tencent own testing.
gpt-5.5,claw_eval_pass3,67.8,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal harness; 20260325 version; 105 queries"",""judge"":""Gemini 3.5 Flash"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass^3"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""agentic/OpenClaw-style""}",False,model_card,verified,Hy3 official model-card appendix: ClawEval (pass^3) = 67.8*. Tencent own testing.
hy3-preview,wildclaw_bench_35_text,45.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""text-only 35-query subset"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""OpenClaw harness""}",True,model_card,verified,"Hy3 official model-card appendix: WildClawBench (35, text-only) = 45.3. Comparator-reported value."
hy3,wildclaw_bench_35_text,53.6,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""text-only 35-query subset"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""OpenClaw harness""}",True,model_card,verified,"Hy3 official model-card appendix: WildClawBench (35, text-only) = 53.6. Comparator-reported value."
glm-5.1,wildclaw_bench_35_text,48.7,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""text-only 35-query subset"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""OpenClaw harness""}",False,model_card,verified,"Hy3 official model-card appendix: WildClawBench (35, text-only) = 48.7*. Tencent own testing."
glm-5.2,wildclaw_bench_35_text,59.2,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""text-only 35-query subset"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""OpenClaw harness""}",False,model_card,verified,"Hy3 official model-card appendix: WildClawBench (35, text-only) = 59.2*. Tencent own testing."
deepseek-v4-flash,wildclaw_bench_35_text,43.5,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""text-only 35-query subset"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""OpenClaw harness""}",False,model_card,verified,"Hy3 official model-card appendix: WildClawBench (35, text-only) = 43.5*. Tencent own testing."
deepseek-v4-pro,wildclaw_bench_35_text,50.5,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""text-only 35-query subset"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""OpenClaw harness""}",False,model_card,verified,"Hy3 official model-card appendix: WildClawBench (35, text-only) = 50.5*. Tencent own testing."
doubao-seed-2.1-pro,wildclaw_bench_35_text,61.7,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""text-only 35-query subset"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""OpenClaw harness""}",False,model_card,verified,"Hy3 official model-card appendix: WildClawBench (35, text-only) = 61.7. Comparator-reported value."
qwen3.7-max,wildclaw_bench_35_text,41.5,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""text-only 35-query subset"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""OpenClaw harness""}",False,model_card,verified,"Hy3 official model-card appendix: WildClawBench (35, text-only) = 41.5*. Tencent own testing."
gemini-3.1-pro,wildclaw_bench_35_text,46.2,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""text-only 35-query subset"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""OpenClaw harness""}",False,model_card,verified,"Hy3 official model-card appendix: WildClawBench (35, text-only) = 46.2*. Tencent own testing."
claude-opus-4.8,wildclaw_bench_35_text,58.4,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""text-only 35-query subset"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""OpenClaw harness""}",False,model_card,verified,"Hy3 official model-card appendix: WildClawBench (35, text-only) = 58.4*. Tencent own testing."
gpt-5.5,wildclaw_bench_35_text,63.2,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""text-only 35-query subset"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""OpenClaw harness""}",False,model_card,verified,"Hy3 official model-card appendix: WildClawBench (35, text-only) = 63.2*. Tencent own testing."
hy3-preview,skills_bench_text_79,29.1,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""79-task self-contained text-only subset; multimodal excluded"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""average over 3 runs"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""Claude Code""}",True,model_card,verified,"Hy3 official model-card appendix: SkillsBench (79, text-only) = 29.1. Comparator-reported value."
hy3,skills_bench_text_79,55.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""79-task self-contained text-only subset; multimodal excluded"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""average over 3 runs"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""Claude Code""}",True,model_card,verified,"Hy3 official model-card appendix: SkillsBench (79, text-only) = 55.3. Comparator-reported value."
glm-5.1,skills_bench_text_79,43.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""79-task self-contained text-only subset; multimodal excluded"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""average over 3 runs"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""Claude Code""}",False,model_card,verified,"Hy3 official model-card appendix: SkillsBench (79, text-only) = 43.0*. Tencent own testing."
glm-5.2,skills_bench_text_79,51.9,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""79-task self-contained text-only subset; multimodal excluded"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""average over 3 runs"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""Claude Code""}",False,model_card,verified,"Hy3 official model-card appendix: SkillsBench (79, text-only) = 51.9*. Tencent own testing."
deepseek-v4-flash,skills_bench_text_79,35.4,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""79-task self-contained text-only subset; multimodal excluded"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""average over 3 runs"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""Claude Code""}",False,model_card,verified,"Hy3 official model-card appendix: SkillsBench (79, text-only) = 35.4*. Tencent own testing."
deepseek-v4-pro,skills_bench_text_79,40.5,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""79-task self-contained text-only subset; multimodal excluded"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""average over 3 runs"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""Claude Code""}",False,model_card,verified,"Hy3 official model-card appendix: SkillsBench (79, text-only) = 40.5*. Tencent own testing."
doubao-seed-2.1-pro,skills_bench_text_79,53.2,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""79-task self-contained text-only subset; multimodal excluded"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""average over 3 runs"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""Claude Code""}",False,model_card,verified,"Hy3 official model-card appendix: SkillsBench (79, text-only) = 53.2*. Tencent own testing."
qwen3.7-max,skills_bench_text_79,59.2,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""79-task self-contained text-only subset; multimodal excluded"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""average over 3 runs"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""Claude Code""}",False,model_card,verified,"Hy3 official model-card appendix: SkillsBench (79, text-only) = 59.2. Comparator-reported value."
claude-opus-4.8,skills_bench_text_79,64.6,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""79-task self-contained text-only subset; multimodal excluded"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""average over 3 runs"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""Claude Code""}",False,model_card,verified,"Hy3 official model-card appendix: SkillsBench (79, text-only) = 64.6*. Tencent own testing."
gpt-5.5,skills_bench_text_79,61.6,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""79-task self-contained text-only subset; multimodal excluded"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""average over 3 runs"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""Claude Code""}",False,model_card,verified,"Hy3 official model-card appendix: SkillsBench (79, text-only) = 61.6*. Tencent own testing."
hy3-preview,e_bench_internal,10.9,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal working-agent benchmark; public protocol not published"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""benchmark-specific agent tools""}",True,model_card,verified,Hy3 official model-card appendix: e-bench (internal) = 10.9. Comparator-reported value.
hy3,e_bench_internal,50.2,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal working-agent benchmark; public protocol not published"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""benchmark-specific agent tools""}",True,model_card,verified,Hy3 official model-card appendix: e-bench (internal) = 50.2. Comparator-reported value.
glm-5.1,e_bench_internal,41.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal working-agent benchmark; public protocol not published"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""benchmark-specific agent tools""}",False,model_card,verified,Hy3 official model-card appendix: e-bench (internal) = 41.0*. Tencent own testing.
glm-5.2,e_bench_internal,52.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal working-agent benchmark; public protocol not published"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""benchmark-specific agent tools""}",False,model_card,verified,Hy3 official model-card appendix: e-bench (internal) = 52.3*. Tencent own testing.
deepseek-v4-flash,e_bench_internal,31.5,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal working-agent benchmark; public protocol not published"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""benchmark-specific agent tools""}",False,model_card,verified,Hy3 official model-card appendix: e-bench (internal) = 31.5*. Tencent own testing.
deepseek-v4-pro,e_bench_internal,34.5,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal working-agent benchmark; public protocol not published"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""benchmark-specific agent tools""}",False,model_card,verified,Hy3 official model-card appendix: e-bench (internal) = 34.5*. Tencent own testing.
doubao-seed-2.1-pro,e_bench_internal,47.9,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal working-agent benchmark; public protocol not published"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""benchmark-specific agent tools""}",False,model_card,verified,Hy3 official model-card appendix: e-bench (internal) = 47.9*. Tencent own testing.
qwen3.7-max,e_bench_internal,50.9,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal working-agent benchmark; public protocol not published"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""benchmark-specific agent tools""}",False,model_card,verified,Hy3 official model-card appendix: e-bench (internal) = 50.9*. Tencent own testing.
gemini-3.1-pro,e_bench_internal,42.6,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal working-agent benchmark; public protocol not published"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""benchmark-specific agent tools""}",False,model_card,verified,Hy3 official model-card appendix: e-bench (internal) = 42.6*. Tencent own testing.
claude-opus-4.8,e_bench_internal,68.8,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal working-agent benchmark; public protocol not published"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""benchmark-specific agent tools""}",False,model_card,verified,Hy3 official model-card appendix: e-bench (internal) = 68.8*. Tencent own testing.
gpt-5.5,e_bench_internal,72.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal working-agent benchmark; public protocol not published"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""benchmark-specific agent tools""}",False,model_card,verified,Hy3 official model-card appendix: e-bench (internal) = 72.0*. Tencent own testing.
hy3-preview,hy_finmodel_bench,26.6,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal working-agent benchmark; public protocol not published"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""benchmark-specific agent tools""}",True,model_card,verified,Hy3 official model-card appendix: Hy-FinModelBench (internal) = 26.6. Comparator-reported value.
hy3,hy_finmodel_bench,69.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal working-agent benchmark; public protocol not published"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""benchmark-specific agent tools""}",True,model_card,verified,Hy3 official model-card appendix: Hy-FinModelBench (internal) = 69.0. Comparator-reported value.
glm-5.1,hy_finmodel_bench,56.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal working-agent benchmark; public protocol not published"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""benchmark-specific agent tools""}",False,model_card,verified,Hy3 official model-card appendix: Hy-FinModelBench (internal) = 56.3*. Tencent own testing.
glm-5.2,hy_finmodel_bench,77.2,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal working-agent benchmark; public protocol not published"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""benchmark-specific agent tools""}",False,model_card,verified,Hy3 official model-card appendix: Hy-FinModelBench (internal) = 77.2*. Tencent own testing.
deepseek-v4-flash,hy_finmodel_bench,54.5,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal working-agent benchmark; public protocol not published"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""benchmark-specific agent tools""}",False,model_card,verified,Hy3 official model-card appendix: Hy-FinModelBench (internal) = 54.5*. Tencent own testing.
deepseek-v4-pro,hy_finmodel_bench,57.6,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal working-agent benchmark; public protocol not published"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""benchmark-specific agent tools""}",False,model_card,verified,Hy3 official model-card appendix: Hy-FinModelBench (internal) = 57.6*. Tencent own testing.
doubao-seed-2.1-pro,hy_finmodel_bench,52.2,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal working-agent benchmark; public protocol not published"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""benchmark-specific agent tools""}",False,model_card,verified,Hy3 official model-card appendix: Hy-FinModelBench (internal) = 52.2*. Tencent own testing.
qwen3.7-max,hy_finmodel_bench,57.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal working-agent benchmark; public protocol not published"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""benchmark-specific agent tools""}",False,model_card,verified,Hy3 official model-card appendix: Hy-FinModelBench (internal) = 57.3*. Tencent own testing.
gemini-3.1-pro,hy_finmodel_bench,54.6,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal working-agent benchmark; public protocol not published"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""benchmark-specific agent tools""}",False,model_card,verified,Hy3 official model-card appendix: Hy-FinModelBench (internal) = 54.6*. Tencent own testing.
claude-opus-4.8,hy_finmodel_bench,74.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal working-agent benchmark; public protocol not published"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""benchmark-specific agent tools""}",False,model_card,verified,Hy3 official model-card appendix: Hy-FinModelBench (internal) = 74.3*. Tencent own testing.
gpt-5.5,hy_finmodel_bench,85.1,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal working-agent benchmark; public protocol not published"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""benchmark-specific agent tools""}",False,model_card,verified,Hy3 official model-card appendix: Hy-FinModelBench (internal) = 85.1*. Tencent own testing.
hy3-preview,prod_bench_internal,10.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""OpenClaw harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass^3"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""OpenClaw harness""}",True,model_card,verified,"Hy3 official model-card appendix: ProdBench (internal, pass^3) = 10.3. Comparator-reported value."
hy3,prod_bench_internal,23.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""OpenClaw harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass^3"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""OpenClaw harness""}",True,model_card,verified,"Hy3 official model-card appendix: ProdBench (internal, pass^3) = 23.0. Comparator-reported value."
glm-5.1,prod_bench_internal,17.5,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""OpenClaw harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass^3"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""OpenClaw harness""}",False,model_card,verified,"Hy3 official model-card appendix: ProdBench (internal, pass^3) = 17.5*. Tencent own testing."
glm-5.2,prod_bench_internal,21.4,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""OpenClaw harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass^3"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""OpenClaw harness""}",False,model_card,verified,"Hy3 official model-card appendix: ProdBench (internal, pass^3) = 21.4*. Tencent own testing."
deepseek-v4-flash,prod_bench_internal,10.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""OpenClaw harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass^3"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""OpenClaw harness""}",False,model_card,verified,"Hy3 official model-card appendix: ProdBench (internal, pass^3) = 10.3*. Tencent own testing."
deepseek-v4-pro,prod_bench_internal,15.1,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""OpenClaw harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass^3"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""OpenClaw harness""}",False,model_card,verified,"Hy3 official model-card appendix: ProdBench (internal, pass^3) = 15.1*. Tencent own testing."
doubao-seed-2.1-pro,prod_bench_internal,18.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""OpenClaw harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass^3"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""OpenClaw harness""}",False,model_card,verified,"Hy3 official model-card appendix: ProdBench (internal, pass^3) = 18.3*. Tencent own testing."
qwen3.7-max,prod_bench_internal,15.9,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""OpenClaw harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass^3"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""OpenClaw harness""}",False,model_card,verified,"Hy3 official model-card appendix: ProdBench (internal, pass^3) = 15.9*. Tencent own testing."
claude-opus-4.8,prod_bench_internal,27.8,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""OpenClaw harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass^3"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""OpenClaw harness""}",False,model_card,verified,"Hy3 official model-card appendix: ProdBench (internal, pass^3) = 27.8*. Tencent own testing."
gpt-5.5,prod_bench_internal,21.4,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""OpenClaw harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass^3"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""OpenClaw harness""}",False,model_card,verified,"Hy3 official model-card appendix: ProdBench (internal, pass^3) = 21.4*. Tencent own testing."
hy3-preview,hy_skillsworld,26.4,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal working-agent benchmark; public protocol not published"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""benchmark-specific agent tools""}",True,model_card,verified,Hy3 official model-card appendix: Hy-SkillsWorld (internal) = 26.4. Comparator-reported value.
hy3,hy_skillsworld,45.8,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal working-agent benchmark; public protocol not published"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""benchmark-specific agent tools""}",True,model_card,verified,Hy3 official model-card appendix: Hy-SkillsWorld (internal) = 45.8. Comparator-reported value.
glm-5.1,hy_skillsworld,37.5,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal working-agent benchmark; public protocol not published"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""benchmark-specific agent tools""}",False,model_card,verified,Hy3 official model-card appendix: Hy-SkillsWorld (internal) = 37.5*. Tencent own testing.
glm-5.2,hy_skillsworld,56.9,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal working-agent benchmark; public protocol not published"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""benchmark-specific agent tools""}",False,model_card,verified,Hy3 official model-card appendix: Hy-SkillsWorld (internal) = 56.9*. Tencent own testing.
deepseek-v4-flash,hy_skillsworld,38.9,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal working-agent benchmark; public protocol not published"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""benchmark-specific agent tools""}",False,model_card,verified,Hy3 official model-card appendix: Hy-SkillsWorld (internal) = 38.9*. Tencent own testing.
deepseek-v4-pro,hy_skillsworld,41.7,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal working-agent benchmark; public protocol not published"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""benchmark-specific agent tools""}",False,model_card,verified,Hy3 official model-card appendix: Hy-SkillsWorld (internal) = 41.7*. Tencent own testing.
doubao-seed-2.1-pro,hy_skillsworld,48.6,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal working-agent benchmark; public protocol not published"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""benchmark-specific agent tools""}",False,model_card,verified,Hy3 official model-card appendix: Hy-SkillsWorld (internal) = 48.6*. Tencent own testing.
qwen3.7-max,hy_skillsworld,42.4,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal working-agent benchmark; public protocol not published"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""benchmark-specific agent tools""}",False,model_card,verified,Hy3 official model-card appendix: Hy-SkillsWorld (internal) = 42.4*. Tencent own testing.
claude-opus-4.8,hy_skillsworld,59.7,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal working-agent benchmark; public protocol not published"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""benchmark-specific agent tools""}",False,model_card,verified,Hy3 official model-card appendix: Hy-SkillsWorld (internal) = 59.7*. Tencent own testing.
gpt-5.5,hy_skillsworld,56.9,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Tencent internal working-agent benchmark; public protocol not published"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""benchmark-specific agent tools""}",False,model_card,verified,Hy3 official model-card appendix: Hy-SkillsWorld (internal) = 56.9*. Tencent own testing.
hy3-preview,hle_tools_text,35.4,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""search/code/web or benchmark-specific tools""}",True,model_card,verified,"Hy3 official model-card appendix: HLE (with tools, text-only) = 35.4. Comparator-reported value."
hy3,hle_tools_text,53.2,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""search/code/web or benchmark-specific tools""}",True,model_card,verified,"Hy3 official model-card appendix: HLE (with tools, text-only) = 53.2. Comparator-reported value."
deepseek-v4-flash,hle_tools_text,45.1,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""search/code/web or benchmark-specific tools""}",False,model_card,verified,"Hy3 official model-card appendix: HLE (with tools, text-only) = 45.1. Comparator-reported value."
doubao-seed-2.1-pro,hle_tools_text,55.7,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""search""}",True,official_blog,verified,Release blog EN: Language/search chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq510uc.png).
gemini-3.1-pro,hle_tools_text,51.4,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""search/code/web or benchmark-specific tools""}",False,model_card,verified,"Hy3 official model-card appendix: HLE (with tools, text-only) = 51.4. Comparator-reported value."
claude-opus-4.8,hle_tools_text,57.9,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""search/code/web or benchmark-specific tools""}",False,model_card,verified,"Hy3 official model-card appendix: HLE (with tools, text-only) = 57.9. Comparator-reported value."
gpt-5.5,hle_tools_text,52.2,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""search/code/web or benchmark-specific tools""}",False,model_card,verified,"Hy3 official model-card appendix: HLE (with tools, text-only) = 52.2. Comparator-reported value."
hy3-preview,hy_euler_pro,2.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""search/code/web or benchmark-specific tools""}",True,model_card,verified,"Hy3 official model-card appendix: Hy-Euler pro (with tools, internal) = 2.3. Comparator-reported value."
hy3,hy_euler_pro,24.2,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""search/code/web or benchmark-specific tools""}",True,model_card,verified,"Hy3 official model-card appendix: Hy-Euler pro (with tools, internal) = 24.2. Comparator-reported value."
glm-5.1,hy_euler_pro,15.9,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""search/code/web or benchmark-specific tools""}",False,model_card,verified,"Hy3 official model-card appendix: Hy-Euler pro (with tools, internal) = 15.9*. Tencent own testing."
glm-5.2,hy_euler_pro,42.4,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""search/code/web or benchmark-specific tools""}",False,model_card,verified,"Hy3 official model-card appendix: Hy-Euler pro (with tools, internal) = 42.4*. Tencent own testing."
deepseek-v4-flash,hy_euler_pro,9.1,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""search/code/web or benchmark-specific tools""}",False,model_card,verified,"Hy3 official model-card appendix: Hy-Euler pro (with tools, internal) = 9.1*. Tencent own testing."
deepseek-v4-pro,hy_euler_pro,25.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""search/code/web or benchmark-specific tools""}",False,model_card,verified,"Hy3 official model-card appendix: Hy-Euler pro (with tools, internal) = 25.0*. Tencent own testing."
doubao-seed-2.1-pro,hy_euler_pro,15.2,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""search/code/web or benchmark-specific tools""}",False,model_card,verified,"Hy3 official model-card appendix: Hy-Euler pro (with tools, internal) = 15.2*. Tencent own testing."
qwen3.7-max,hy_euler_pro,50.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""search/code/web or benchmark-specific tools""}",False,model_card,verified,"Hy3 official model-card appendix: Hy-Euler pro (with tools, internal) = 50.0*. Tencent own testing."
gemini-3.1-pro,hy_euler_pro,54.6,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""search/code/web or benchmark-specific tools""}",False,model_card,verified,"Hy3 official model-card appendix: Hy-Euler pro (with tools, internal) = 54.6*. Tencent own testing."
claude-opus-4.8,hy_euler_pro,64.4,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""search/code/web or benchmark-specific tools""}",False,model_card,verified,"Hy3 official model-card appendix: Hy-Euler pro (with tools, internal) = 64.4*. Tencent own testing."
gpt-5.5,hy_euler_pro,77.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""search/code/web or benchmark-specific tools""}",False,model_card,verified,"Hy3 official model-card appendix: Hy-Euler pro (with tools, internal) = 77.3*. Tencent own testing."
hy3-preview,gpqa_diamond,87.2,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",True,model_card,verified,Hy3 official model-card appendix: GPQA Diamond = 87.2. Comparator-reported value.
hy3,gpqa_diamond,90.4,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",True,model_card,verified,Hy3 official model-card appendix: GPQA Diamond = 90.4. Comparator-reported value.
doubao-seed-2.1-pro,gpqa_diamond,89.8,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: GPQA Diamond = 89.8*. Tencent own testing.
hy3-preview,hle_text,30.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",True,model_card,verified,"Hy3 official model-card appendix: HLE (no tools, text-only) = 30.0. Comparator-reported value."
hy3,hle_text,37.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",True,model_card,verified,"Hy3 official model-card appendix: HLE (no tools, text-only) = 37.0. Comparator-reported value."
doubao-seed-2.1-pro,hle_text,35.9,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,"Hy3 official model-card appendix: HLE (no tools, text-only) = 35.9*. Tencent own testing."
claude-opus-4.8,hle_text,49.8,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,"Hy3 official model-card appendix: HLE (no tools, text-only) = 49.8. Comparator-reported value."
hy3-preview,frontier_science_research,19.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",True,model_card,verified,Hy3 official model-card appendix: FrontierScience-Research = 19.0. Comparator-reported value.
hy3,frontier_science_research,21.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",True,model_card,verified,Hy3 official model-card appendix: FrontierScience-Research = 21.3. Comparator-reported value.
glm-5.1,frontier_science_research,13.5,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: FrontierScience-Research = 13.5*. Tencent own testing.
glm-5.2,frontier_science_research,19.8,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: FrontierScience-Research = 19.8*. Tencent own testing.
deepseek-v4-flash,frontier_science_research,16.7,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: FrontierScience-Research = 16.7*. Tencent own testing.
deepseek-v4-pro,frontier_science_research,21.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: FrontierScience-Research = 21.3*. Tencent own testing.
doubao-seed-2.1-pro,frontier_science_research,28.3,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""research-oriented agent environment""}",True,official_blog,verified,Release blog EN: Frontier-research chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq52cdr.png).
qwen3.7-max,frontier_science_research,24.8,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: FrontierScience-Research = 24.8*. Tencent own testing.
gemini-3.1-pro,frontier_science_research,16.7,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: FrontierScience-Research = 16.7. Comparator-reported value.
claude-opus-4.8,frontier_science_research,32.9,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: FrontierScience-Research = 32.9*. Tencent own testing.
gpt-5.5,frontier_science_research,33.9,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: FrontierScience-Research = 33.9. Comparator-reported value.
hy3-preview,frontiersci_olympiad,70.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official paper judge prompts; self-evaluated"",""judge"":""gpt-oss-120b, high reasoning"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",True,model_card,verified,Hy3 official model-card appendix: FrontierScience-Olympiad = 70.0. Comparator-reported value.
hy3,frontiersci_olympiad,74.8,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official paper judge prompts; self-evaluated"",""judge"":""gpt-oss-120b, high reasoning"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",True,model_card,verified,Hy3 official model-card appendix: FrontierScience-Olympiad = 74.8. Comparator-reported value.
glm-5.1,frontiersci_olympiad,65.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official paper judge prompts; self-evaluated"",""judge"":""gpt-oss-120b, high reasoning"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: FrontierScience-Olympiad = 65.0*. Tencent own testing.
glm-5.2,frontiersci_olympiad,72.5,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official paper judge prompts; self-evaluated"",""judge"":""gpt-oss-120b, high reasoning"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: FrontierScience-Olympiad = 72.5*. Tencent own testing.
deepseek-v4-flash,frontiersci_olympiad,73.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official paper judge prompts; self-evaluated"",""judge"":""gpt-oss-120b, high reasoning"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: FrontierScience-Olympiad = 73.0*. Tencent own testing.
deepseek-v4-pro,frontiersci_olympiad,70.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official paper judge prompts; self-evaluated"",""judge"":""gpt-oss-120b, high reasoning"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: FrontierScience-Olympiad = 70.0*. Tencent own testing.
doubao-seed-2.1-pro,frontiersci_olympiad,75.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Language/search chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq510uc.png).
qwen3.7-max,frontiersci_olympiad,74.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official paper judge prompts; self-evaluated"",""judge"":""gpt-oss-120b, high reasoning"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: FrontierScience-Olympiad = 74.3*. Tencent own testing.
gemini-3.1-pro,frontiersci_olympiad,79.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official paper judge prompts; self-evaluated"",""judge"":""gpt-oss-120b, high reasoning"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: FrontierScience-Olympiad = 79.0. Comparator-reported value.
claude-opus-4.8,frontiersci_olympiad,74.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official paper judge prompts; self-evaluated"",""judge"":""gpt-oss-120b, high reasoning"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: FrontierScience-Olympiad = 74.3*. Tencent own testing.
gpt-5.5,frontiersci_olympiad,73.8,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official paper judge prompts; self-evaluated"",""judge"":""gpt-oss-120b, high reasoning"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: FrontierScience-Olympiad = 73.8*. Tencent own testing.
hy3-preview,usamo_2026,37.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",True,model_card,verified,Hy3 official model-card appendix: USAMO 2026 = 37.3. Comparator-reported value.
hy3,usamo_2026,72.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",True,model_card,verified,Hy3 official model-card appendix: USAMO 2026 = 72.0. Comparator-reported value.
glm-5.1,usamo_2026,37.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: USAMO 2026 = 37.3*. Tencent own testing.
glm-5.2,usamo_2026,41.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: USAMO 2026 = 41.3*. Tencent own testing.
deepseek-v4-flash,usamo_2026,57.9,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: USAMO 2026 = 57.9*. Tencent own testing.
deepseek-v4-pro,usamo_2026,59.2,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: USAMO 2026 = 59.2*. Tencent own testing.
doubao-seed-2.1-pro,usamo_2026,60.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: USAMO 2026 = 60.3*. Tencent own testing.
qwen3.7-max,usamo_2026,57.1,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: USAMO 2026 = 57.1*. Tencent own testing.
claude-opus-4.8,usamo_2026,96.7,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""high"",""harness"":""usamo"",""sampling"":""10 attempts/problem"",""tools"":""none""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Section 8.6, page 120: Claude Opus 4.8; USAMO 2026 [summary]=96.7%. Source setting: effort=high; tools=none; sampling=10 attempts/problem; harness=usamo."
gpt-5.5,usamo_2026,98.4,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: USAMO 2026 = 98.4*. Tencent own testing.
hy3-preview,matharena_apex_2025,12.6,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""MathArena Apex 2025, 12-problem set"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""samples=16 independent runs per problem"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",True,model_card,verified,Hy3 official model-card appendix: MathArena Apex = 12.6. Comparator-reported value.
hy3,matharena_apex_2025,38.7,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""MathArena Apex 2025, 12-problem set"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""samples=16 independent runs per problem"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",True,model_card,verified,Hy3 official model-card appendix: MathArena Apex = 38.7. Comparator-reported value.
glm-5.2,matharena_apex_2025,16.8,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""MathArena Apex 2025, 12-problem set"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""samples=16 independent runs per problem"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: MathArena Apex = 16.8*. Tencent own testing.
doubao-seed-2.1-pro,matharena_apex_2025,31.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 11 (p53). Source dash is not zero.
qwen3.7-max,matharena_apex_2025,44.5,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""MathArena Apex 2025, 12-problem set"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""samples=16 independent runs per problem"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: MathArena Apex = 44.5. Comparator-reported value.
gpt-5.5,matharena_apex_2025,85.4,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""MathArena Apex 2025, 12-problem set"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""samples=16 independent runs per problem"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: MathArena Apex = 85.4*. Tencent own testing.
hy3-preview,horizon_math_pass12,1.8,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@12"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",True,model_card,verified,Hy3 official model-card appendix: HorizonMath (pass@12) = 1.8. Comparator-reported value.
hy3,horizon_math_pass12,7.1,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@12"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",True,model_card,verified,Hy3 official model-card appendix: HorizonMath (pass@12) = 7.1. Comparator-reported value.
glm-5.2,horizon_math_pass12,7.1,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@12"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: HorizonMath (pass@12) = 7.1*. Tencent own testing.
deepseek-v4-flash,horizon_math_pass12,4.4,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@12"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: HorizonMath (pass@12) = 4.4*. Tencent own testing.
deepseek-v4-pro,horizon_math_pass12,5.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@12"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: HorizonMath (pass@12) = 5.3*. Tencent own testing.
doubao-seed-2.1-pro,horizon_math_pass12,2.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""research-oriented agent environment""}",True,official_blog,verified,Release blog EN: Frontier-research chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq52cdr.png).
qwen3.7-max,horizon_math_pass12,6.2,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@12"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: HorizonMath (pass@12) = 6.2*. Tencent own testing.
gemini-3.1-pro,horizon_math_pass12,7.1,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@12"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: HorizonMath (pass@12) = 7.1*. Tencent own testing.
gpt-5.5,horizon_math_pass12,11.6,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@12"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: HorizonMath (pass@12) = 11.6*. Tencent own testing.
hy3-preview,hy_math_internal,26.1,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",True,model_card,verified,Hy3 official model-card appendix: Hy-Math (internal) = 26.1. Comparator-reported value.
hy3,hy_math_internal,60.9,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",True,model_card,verified,Hy3 official model-card appendix: Hy-Math (internal) = 60.9. Comparator-reported value.
glm-5.1,hy_math_internal,6.7,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: Hy-Math (internal) = 6.7*. Tencent own testing.
glm-5.2,hy_math_internal,16.4,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: Hy-Math (internal) = 16.4*. Tencent own testing.
deepseek-v4-flash,hy_math_internal,66.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: Hy-Math (internal) = 66.0*. Tencent own testing.
deepseek-v4-pro,hy_math_internal,63.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: Hy-Math (internal) = 63.0*. Tencent own testing.
doubao-seed-2.1-pro,hy_math_internal,65.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: Hy-Math (internal) = 65.3*. Tencent own testing.
qwen3.7-max,hy_math_internal,66.1,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: Hy-Math (internal) = 66.1*. Tencent own testing.
gemini-3.1-pro,hy_math_internal,63.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: Hy-Math (internal) = 63.3*. Tencent own testing.
gpt-5.5,hy_math_internal,91.4,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: Hy-Math (internal) = 91.4*. Tencent own testing.
hy3-preview,phybench,71.4,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",True,model_card,verified,Hy3 official model-card appendix: PHYBench = 71.4. Comparator-reported value.
hy3,phybench,77.4,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",True,model_card,verified,Hy3 official model-card appendix: PHYBench = 77.4. Comparator-reported value.
glm-5.1,phybench,70.9,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: PHYBench = 70.9*. Tencent own testing.
glm-5.2,phybench,71.5,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: PHYBench = 71.5*. Tencent own testing.
deepseek-v4-flash,phybench,76.6,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: PHYBench = 76.6*. Tencent own testing.
deepseek-v4-pro,phybench,75.4,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: PHYBench = 75.4*. Tencent own testing.
doubao-seed-2.1-pro,phybench,77.5,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: PHYBench = 77.5*. Tencent own testing.
qwen3.7-max,phybench,76.5,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: PHYBench = 76.5*. Tencent own testing.
gemini-3.1-pro,phybench,82.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: PHYBench = 82.0*. Tencent own testing.
claude-opus-4.8,phybench,79.6,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: PHYBench = 79.6*. Tencent own testing.
gpt-5.5,phybench,77.5,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: PHYBench = 77.5*. Tencent own testing.
hy3-preview,cmt_benchmark,19.4,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",True,model_card,verified,Hy3 official model-card appendix: CMT-Benchmark = 19.4. Comparator-reported value.
hy3,cmt_benchmark,37.8,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",True,model_card,verified,Hy3 official model-card appendix: CMT-Benchmark = 37.8. Comparator-reported value.
glm-5.1,cmt_benchmark,25.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: CMT-Benchmark = 25.3*. Tencent own testing.
glm-5.2,cmt_benchmark,34.1,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: CMT-Benchmark = 34.1*. Tencent own testing.
deepseek-v4-flash,cmt_benchmark,26.8,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: CMT-Benchmark = 26.8*. Tencent own testing.
deepseek-v4-pro,cmt_benchmark,35.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: CMT-Benchmark = 35.0*. Tencent own testing.
doubao-seed-2.1-pro,cmt_benchmark,31.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: CMT-Benchmark = 31.0*. Tencent own testing.
qwen3.7-max,cmt_benchmark,40.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: CMT-Benchmark = 40.0*. Tencent own testing.
gemini-3.1-pro,cmt_benchmark,47.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: CMT-Benchmark = 47.0*. Tencent own testing.
claude-opus-4.8,cmt_benchmark,43.8,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: CMT-Benchmark = 43.8*. Tencent own testing.
gpt-5.5,cmt_benchmark,43.6,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: CMT-Benchmark = 43.6*. Tencent own testing.
hy3-preview,imo_answerbench,84.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",True,model_card,verified,Hy3 official model-card appendix: IMOAnswerBench = 84.3. Comparator-reported value.
hy3,imo_answerbench,90.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",True,model_card,verified,Hy3 official model-card appendix: IMOAnswerBench = 90.0. Comparator-reported value.
doubao-seed-2.1-pro,imo_answerbench,88.5,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: IMOAnswerBench = 88.5*. Tencent own testing.
gpt-5.5,imo_answerbench,92.1,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: IMOAnswerBench = 92.1*. Tencent own testing.
hy3-preview,superchem,47.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",True,model_card,verified,Hy3 official model-card appendix: SuperChem = 47.0. Comparator-reported value.
hy3,superchem,54.9,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",True,model_card,verified,Hy3 official model-card appendix: SuperChem = 54.9. Comparator-reported value.
glm-5.1,superchem,51.4,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: SuperChem = 51.4*. Tencent own testing.
glm-5.2,superchem,60.2,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: SuperChem = 60.2*. Tencent own testing.
deepseek-v4-flash,superchem,51.8,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: SuperChem = 51.8*. Tencent own testing.
deepseek-v4-pro,superchem,61.2,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: SuperChem = 61.2*. Tencent own testing.
doubao-seed-2.1-pro,superchem,59.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 11 (p53).
qwen3.7-max,superchem,60.5,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: SuperChem = 60.5*. Tencent own testing.
gemini-3.1-pro,superchem,69.2,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: SuperChem = 69.2*. Tencent own testing.
claude-opus-4.8,superchem,66.9,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: SuperChem = 66.9*. Tencent own testing.
gpt-5.5,superchem,66.2,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: SuperChem = 66.2*. Tencent own testing.
hy3-preview,cl_bench,22.8,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",True,model_card,verified,Hy3 official model-card appendix: CL-bench = 22.8. Comparator-reported value.
hy3,cl_bench,23.8,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",True,model_card,verified,Hy3 official model-card appendix: CL-bench = 23.8. Comparator-reported value.
glm-5.1,cl_bench,21.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: CL-bench = 21.0*. Tencent own testing.
glm-5.2,cl_bench,23.1,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: CL-bench = 23.1*. Tencent own testing.
deepseek-v4-flash,cl_bench,16.4,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: CL-bench = 16.4*. Tencent own testing.
deepseek-v4-pro,cl_bench,18.1,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: CL-bench = 18.1*. Tencent own testing.
doubao-seed-2.1-pro,cl_bench,23.5,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: CL-bench = 23.5*. Tencent own testing.
qwen3.7-max,cl_bench,18.4,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: CL-bench = 18.4*. Tencent own testing.
gemini-3.1-pro,cl_bench,21.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: CL-bench = 21.3*. Tencent own testing.
claude-opus-4.8,cl_bench,24.8,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: CL-bench = 24.8*. Tencent own testing.
gpt-5.5,cl_bench,27.8,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: CL-bench = 27.8*. Tencent own testing.
hy3-preview,cl_bench_life,15.7,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",True,model_card,verified,Hy3 official model-card appendix: CL-bench life = 15.7. Comparator-reported value.
hy3,cl_bench_life,17.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",True,model_card,verified,Hy3 official model-card appendix: CL-bench life = 17.0. Comparator-reported value.
glm-5.1,cl_bench_life,14.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: CL-bench life = 14.0*. Tencent own testing.
glm-5.2,cl_bench_life,21.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: CL-bench life = 21.0*. Tencent own testing.
deepseek-v4-flash,cl_bench_life,11.4,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: CL-bench life = 11.4*. Tencent own testing.
deepseek-v4-pro,cl_bench_life,14.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: CL-bench life = 14.3*. Tencent own testing.
doubao-seed-2.1-pro,cl_bench_life,13.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: CL-bench life = 13.3*. Tencent own testing.
qwen3.7-max,cl_bench_life,13.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: CL-bench life = 13.0*. Tencent own testing.
gemini-3.1-pro,cl_bench_life,16.5,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: CL-bench life = 16.5*. Tencent own testing.
claude-opus-4.8,cl_bench_life,16.9,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: CL-bench life = 16.9*. Tencent own testing.
gpt-5.5,cl_bench_life,21.1,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""official benchmark or reported provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: CL-bench life = 21.1*. Tencent own testing.
hy3-preview,aa_lcr,66.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Artificial Analysis AA-LCR, 100 questions"",""judge"":""Equality Checker LLM"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""repeats=3 per question"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",True,model_card,verified,Hy3 official model-card appendix: AA-LCR = 66.3. Comparator-reported value.
hy3,aa_lcr,73.4,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Artificial Analysis AA-LCR, 100 questions"",""judge"":""Equality Checker LLM"",""mode"":""thinking"",""notes"":""Comparator value not marked as Tencent own test"",""prompt_style"":""default"",""sampling"":""repeats=3 per question"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",True,model_card,verified,Hy3 official model-card appendix: AA-LCR = 73.4. Comparator-reported value.
glm-5.1,aa_lcr,71.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Artificial Analysis AA-LCR, 100 questions"",""judge"":""Equality Checker LLM"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""repeats=3 per question"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: AA-LCR = 71.0*. Tencent own testing.
deepseek-v4-flash,aa_lcr,64.8,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Artificial Analysis AA-LCR, 100 questions"",""judge"":""Equality Checker LLM"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""repeats=3 per question"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: AA-LCR = 64.8*. Tencent own testing.
deepseek-v4-pro,aa_lcr,71.3,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Artificial Analysis AA-LCR, 100 questions"",""judge"":""Equality Checker LLM"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""repeats=3 per question"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: AA-LCR = 71.3*. Tencent own testing.
doubao-seed-2.1-pro,aa_lcr,70.0,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Artificial Analysis AA-LCR, 100 questions"",""judge"":""Equality Checker LLM"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""repeats=3 per question"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: AA-LCR = 70.0*. Tencent own testing.
qwen3.7-max,aa_lcr,70.2,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Artificial Analysis AA-LCR, 100 questions"",""judge"":""Equality Checker LLM"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""repeats=3 per question"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: AA-LCR = 70.2*. Tencent own testing.
gemini-3.1-pro,aa_lcr,74.6,https://huggingface.co/tencent/Hy3,"{""context"":""256000 for Hy3 family; otherwise source/provider setting"",""effort"":""high / highest available"",""harness"":""Artificial Analysis AA-LCR, 100 questions"",""judge"":""Equality Checker LLM"",""mode"":""thinking"",""notes"":""Tencent own test"",""prompt_style"":""default"",""sampling"":""repeats=3 per question"",""temperature"":""0.9 for Hy3 family; otherwise source/provider setting"",""tools"":""none""}",False,model_card,verified,Hy3 official model-card appendix: AA-LCR = 74.6*. Tencent own testing.
longcat-2.0,terminal_bench_2_1,70.8,https://longcat.chat/blog/longcat-2.0/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""Claude Code; 8 CPU/16GB; timeout=6h"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Measured in-house by Meituan unless otherwise noted."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_k=-1; top_p=0.95"",""tools"":""terminal/code execution""}",True,official_blog,verified,LongCat-2.0 official tech blog: Terminal-Bench 2.1 = 70.8. Measured in-house by Meituan under the reported unified harness.
claude-opus-4.7,terminal_bench_2_1,71.7,https://longcat.chat/blog/longcat-2.0/,"{""context"":""default"",""effort"":""default"",""harness"":""Terminus-2"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""notes"":""Asterisked provider-reported value. Opus 4.7 canonical Terminal-Bench setting uses Terminus-2 with thinking disabled."",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 attempts)"",""temperature"":""provider official setting"",""tools"":""terminal agent""}",True,official_blog,verified,LongCat-2.0 official tech blog: Terminal-Bench 2.1 = 71.7*. Asterisked value is cited by LongCat from the model provider's official report.
longcat-2.0,swe_bench_pro,59.51,https://longcat.chat/blog/longcat-2.0/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""Claude Code; 4 CPU/8GB; problematic tasks corrected"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Measured in-house by Meituan unless otherwise noted."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_k=-1; top_p=1.0"",""tools"":""agentic code execution""}",True,official_blog,verified,LongCat-2.0 official tech blog: SWE-bench Pro = 59.5 (chart data 59.51). Measured in-house by Meituan under the reported unified harness.
longcat-2.0,swe_bench_multilingual,77.33,https://longcat.chat/blog/longcat-2.0/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""Claude Code; 4 CPU/8GB; problematic tasks corrected"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Measured in-house by Meituan unless otherwise noted."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""1.0; top_k=-1; top_p=1.0"",""tools"":""agentic code execution""}",True,official_blog,verified,LongCat-2.0 official tech blog: SWE-bench Multilingual = 77.3 (chart data 77.33). Measured in-house by Meituan under the reported unified harness.
claude-opus-4.7,swe_bench_multilingual,80.5,https://longcat.chat/blog/longcat-2.0/,"{""context"":""default"",""effort"":""default"",""harness"":""provider official report; details not reproduced by LongCat"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Asterisk: LongCat cites this value from the model provider's official report rather than measuring it in-house."",""prompt_style"":""default"",""sampling"":""pass@1 (avg 5 trials)"",""temperature"":""provider official setting"",""tools"":""agentic code execution""}",True,official_blog,verified,LongCat-2.0 official tech blog: SWE-bench Multilingual = 80.5*. Asterisked value is cited by LongCat from the model provider's official report.
longcat-2.0,forte_avg3,73.22,https://longcat.chat/blog/longcat-2.0/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""OpenClaw in Docker; timeout=45m; 2 CPU/4GB; single API call=500s; max retries=10"",""judge"":""LLM-as-judge; all-or-nothing rubric score"",""mode"":""thinking"",""notes"":""Measured in-house by Meituan unless otherwise noted."",""prompt_style"":""default"",""sampling"":""trials=3 (Avg@3)"",""temperature"":""source does not state"",""tools"":""office-computing environment""}",True,official_blog,verified,LongCat-2.0 official tech blog: FORTE = 73.2 (chart data 73.22). Measured in-house by Meituan under the reported unified harness.
gemini-3.1-pro,forte_avg3,70.31,https://longcat.chat/blog/longcat-2.0/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""OpenClaw in Docker; timeout=45m; 2 CPU/4GB; single API call=500s; max retries=10"",""judge"":""LLM-as-judge; all-or-nothing rubric score"",""mode"":""thinking"",""notes"":""Measured in-house by Meituan unless otherwise noted."",""prompt_style"":""default"",""sampling"":""trials=3 (Avg@3)"",""temperature"":""source does not state"",""tools"":""office-computing environment""}",False,official_blog,verified,LongCat-2.0 official tech blog: FORTE = 70.3 (chart data 70.31). Measured in-house by Meituan under the reported unified harness.
gpt-5.5,forte_avg3,77.78,https://longcat.chat/blog/longcat-2.0/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""OpenClaw in Docker; timeout=45m; 2 CPU/4GB; single API call=500s; max retries=10"",""judge"":""LLM-as-judge; all-or-nothing rubric score"",""mode"":""thinking"",""notes"":""Measured in-house by Meituan unless otherwise noted."",""prompt_style"":""default"",""sampling"":""trials=3 (Avg@3)"",""temperature"":""source does not state"",""tools"":""office-computing environment""}",False,official_blog,verified,LongCat-2.0 official tech blog: FORTE = 77.8 (chart data 77.78). Measured in-house by Meituan under the reported unified harness.
claude-opus-4.6,forte_avg3,73.22,https://longcat.chat/blog/longcat-2.0/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""OpenClaw in Docker; timeout=45m; 2 CPU/4GB; single API call=500s; max retries=10"",""judge"":""LLM-as-judge; all-or-nothing rubric score"",""mode"":""thinking"",""notes"":""Measured in-house by Meituan unless otherwise noted."",""prompt_style"":""default"",""sampling"":""trials=3 (Avg@3)"",""temperature"":""source does not state"",""tools"":""office-computing environment""}",False,official_blog,verified,LongCat-2.0 official tech blog: FORTE = 73.2 (chart data 73.22). Measured in-house by Meituan under the reported unified harness.
claude-opus-4.7,forte_avg3,77.6,https://longcat.chat/blog/longcat-2.0/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""OpenClaw in Docker; timeout=45m; 2 CPU/4GB; single API call=500s; max retries=10"",""judge"":""LLM-as-judge; all-or-nothing rubric score"",""mode"":""thinking"",""notes"":""Measured in-house by Meituan unless otherwise noted."",""prompt_style"":""default"",""sampling"":""trials=3 (Avg@3)"",""temperature"":""source does not state"",""tools"":""office-computing environment""}",False,official_blog,verified,LongCat-2.0 official tech blog: FORTE = 77.6. Measured in-house by Meituan under the reported unified harness.
claude-opus-4.8,forte_avg3,77.23,https://longcat.chat/blog/longcat-2.0/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""OpenClaw in Docker; timeout=45m; 2 CPU/4GB; single API call=500s; max retries=10"",""judge"":""LLM-as-judge; all-or-nothing rubric score"",""mode"":""thinking"",""notes"":""Measured in-house by Meituan unless otherwise noted."",""prompt_style"":""default"",""sampling"":""trials=3 (Avg@3)"",""temperature"":""source does not state"",""tools"":""office-computing environment""}",False,official_blog,verified,LongCat-2.0 official tech blog: FORTE = 77.2 (chart data 77.23). Measured in-house by Meituan under the reported unified harness.
longcat-2.0,browsecomp,79.86,https://longcat.chat/blog/longcat-2.0/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""Meituan unified in-house search harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Measured in-house by Meituan unless otherwise noted."",""prompt_style"":""default"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""Search and Browse""}",True,official_blog,verified,LongCat-2.0 official tech blog: BrowseComp = 79.9 (chart data 79.86). Measured in-house by Meituan under the reported unified harness.
longcat-2.0,rwsearch,78.75,https://longcat.chat/blog/longcat-2.0/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""bare-model evaluation; no context management"",""judge"":""publisher objective scorer; details undisclosed"",""mode"":""thinking"",""notes"":""Measured in-house by Meituan unless otherwise noted."",""prompt_style"":""default"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""basic Search and Browse""}",True,official_blog,verified,LongCat-2.0 official tech blog: RWSearch = 78.8 (chart data 78.75). Measured in-house by Meituan under the reported unified harness.
gemini-3.1-pro,rwsearch,76.25,https://longcat.chat/blog/longcat-2.0/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""bare-model evaluation; no context management"",""judge"":""publisher objective scorer; details undisclosed"",""mode"":""thinking"",""notes"":""Measured in-house by Meituan unless otherwise noted."",""prompt_style"":""default"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""basic Search and Browse""}",False,official_blog,verified,LongCat-2.0 official tech blog: RWSearch = 76.3 (chart data 76.25). Measured in-house by Meituan under the reported unified harness.
gpt-5.5,rwsearch,85.25,https://longcat.chat/blog/longcat-2.0/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""bare-model evaluation; no context management"",""judge"":""publisher objective scorer; details undisclosed"",""mode"":""thinking"",""notes"":""Measured in-house by Meituan unless otherwise noted."",""prompt_style"":""default"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""basic Search and Browse""}",False,official_blog,verified,LongCat-2.0 official tech blog: RWSearch = 85.3 (chart data 85.25). Measured in-house by Meituan under the reported unified harness.
claude-opus-4.6,rwsearch,81.25,https://longcat.chat/blog/longcat-2.0/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""bare-model evaluation; no context management"",""judge"":""publisher objective scorer; details undisclosed"",""mode"":""thinking"",""notes"":""Measured in-house by Meituan unless otherwise noted."",""prompt_style"":""default"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""basic Search and Browse""}",False,official_blog,verified,LongCat-2.0 official tech blog: RWSearch = 81.3 (chart data 81.25). Measured in-house by Meituan under the reported unified harness.
claude-opus-4.7,rwsearch,79.25,https://longcat.chat/blog/longcat-2.0/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""bare-model evaluation; no context management"",""judge"":""publisher objective scorer; details undisclosed"",""mode"":""thinking"",""notes"":""Measured in-house by Meituan unless otherwise noted."",""prompt_style"":""default"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""basic Search and Browse""}",False,official_blog,verified,LongCat-2.0 official tech blog: RWSearch = 79.3 (chart data 79.25). Measured in-house by Meituan under the reported unified harness.
claude-opus-4.8,rwsearch,77.25,https://longcat.chat/blog/longcat-2.0/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""bare-model evaluation; no context management"",""judge"":""publisher objective scorer; details undisclosed"",""mode"":""thinking"",""notes"":""Measured in-house by Meituan unless otherwise noted."",""prompt_style"":""default"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""basic Search and Browse""}",False,official_blog,verified,LongCat-2.0 official tech blog: RWSearch = 77.3 (chart data 77.25). Measured in-house by Meituan under the reported unified harness.
longcat-2.0,ifeval,90.0,https://longcat.chat/blog/longcat-2.0/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""Meituan unified in-house harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Measured in-house by Meituan unless otherwise noted."",""prompt_style"":""default"",""sampling"":""source does not state"",""temperature"":""0.7; top_k=-1; top_p=0.95"",""tools"":""none""}",True,official_blog,verified,LongCat-2.0 official tech blog: IFEval = 90.0. Measured in-house by Meituan under the reported unified harness.
gpt-5.5,ifeval,95.0,https://longcat.chat/blog/longcat-2.0/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""Meituan unified in-house harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Measured in-house by Meituan unless otherwise noted."",""prompt_style"":""default"",""sampling"":""source does not state"",""temperature"":""0.7; top_k=-1; top_p=0.95"",""tools"":""none""}",False,official_blog,verified,LongCat-2.0 official tech blog: IFEval = 95.0. Measured in-house by Meituan under the reported unified harness.
claude-opus-4.7,ifeval,88.7,https://longcat.chat/blog/longcat-2.0/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""Meituan unified in-house harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Measured in-house by Meituan unless otherwise noted."",""prompt_style"":""default"",""sampling"":""source does not state"",""temperature"":""0.7; top_k=-1; top_p=0.95"",""tools"":""none""}",False,official_blog,verified,LongCat-2.0 official tech blog: IFEval = 88.7. Measured in-house by Meituan under the reported unified harness.
claude-opus-4.8,ifeval,86.0,https://longcat.chat/blog/longcat-2.0/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""Meituan unified in-house harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Measured in-house by Meituan unless otherwise noted."",""prompt_style"":""default"",""sampling"":""source does not state"",""temperature"":""0.7; top_k=-1; top_p=0.95"",""tools"":""none""}",False,official_blog,verified,LongCat-2.0 official tech blog: IFEval = 86.0. Measured in-house by Meituan under the reported unified harness.
longcat-2.0,imo_answerbench,81.8,https://longcat.chat/blog/longcat-2.0/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""Meituan unified in-house harness"",""judge"":""IMO-AnswerBench answer autograder"",""mode"":""thinking"",""notes"":""Measured in-house by Meituan unless otherwise noted."",""prompt_style"":""default"",""sampling"":""source does not state"",""temperature"":""1.0; top_k=-1; top_p=0.95"",""tools"":""none""}",True,official_blog,verified,LongCat-2.0 official tech blog: IMO-AnswerBench = 81.8. Measured in-house by Meituan under the reported unified harness.
claude-opus-4.7,imo_answerbench,81.8,https://longcat.chat/blog/longcat-2.0/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""Meituan unified in-house harness"",""judge"":""IMO-AnswerBench answer autograder"",""mode"":""thinking"",""notes"":""Measured in-house by Meituan unless otherwise noted."",""prompt_style"":""default"",""sampling"":""source does not state"",""temperature"":""1.0; top_k=-1; top_p=0.95"",""tools"":""none""}",False,official_blog,verified,LongCat-2.0 official tech blog: IMO-AnswerBench = 81.8. Measured in-house by Meituan under the reported unified harness.
longcat-2.0,gpqa_diamond,88.9,https://longcat.chat/blog/longcat-2.0/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""Meituan unified in-house harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Measured in-house by Meituan unless otherwise noted."",""prompt_style"":""default"",""sampling"":""source does not state"",""temperature"":""0.7; top_k=-1; top_p=0.95"",""tools"":""none""}",True,official_blog,verified,LongCat-2.0 official tech blog: GPQA-diamond = 88.9. Measured in-house by Meituan under the reported unified harness.
intern-s2-preview-397b,mmlu_pro,89.75,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""max 256K text"",""effort"":""source does not state"",""harness"":""OpenCompass or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""thinking (default)"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",True,model_card,verified,Intern-S2-Preview-397B official model card: MMLU Pro = 89.75. This is the release model's own reported evaluation.
kimi-k2.7-code,mmlu_pro,87.1,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""OpenCompass or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: MMLU Pro = 87.1. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
glm-5.2,mmlu_pro,87.22,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""OpenCompass or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: MMLU Pro = 87.22. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
gpt-5.5,mmlu_pro,88.2,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""OpenCompass or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: MMLU Pro = 88.2. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
claude-opus-4.8,mmlu_pro,90.12,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""OpenCompass or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: MMLU Pro = 90.12. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
intern-s2-preview-397b,simpleqa_verified,69.9,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""max 256K text"",""effort"":""source does not state"",""harness"":""OpenCompass or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""thinking (default)"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",True,model_card,verified,Intern-S2-Preview-397B official model card: SimpleQA-Verified = 69.9. This is the release model's own reported evaluation.
qwen3.5-397b,simpleqa_verified,54.8,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""OpenCompass or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: SimpleQA-Verified = 54.8. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
kimi-k2.7-code,simpleqa_verified,38.6,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""OpenCompass or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: SimpleQA-Verified = 38.6. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
glm-5.2,simpleqa_verified,37.9,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""OpenCompass or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: SimpleQA-Verified = 37.9. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
gpt-5.5,simpleqa_verified,64.3,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""OpenCompass or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: SimpleQA-Verified = 64.3. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
claude-opus-4.8,simpleqa_verified,43.3,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""OpenCompass or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: SimpleQA-Verified = 43.3. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
intern-s2-preview-397b,advancedif,74.44,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""max 256K text"",""effort"":""source does not state"",""harness"":""OpenCompass or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""thinking (default)"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",True,model_card,verified,Intern-S2-Preview-397B official model card: AdvancedIF = 74.44. This is the release model's own reported evaluation.
qwen3.5-397b,advancedif,75.49,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""OpenCompass or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: AdvancedIF = 75.49. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
deepseek-v4-pro,advancedif,73.83,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""OpenCompass or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: AdvancedIF = 73.83. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
kimi-k2.7-code,advancedif,76.17,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""OpenCompass or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: AdvancedIF = 76.17. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
glm-5.2,advancedif,75.76,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""OpenCompass or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: AdvancedIF = 75.76. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
gpt-5.5,advancedif,76.2,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""OpenCompass or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: AdvancedIF = 76.2. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
gemini-3.1-pro,advancedif,79.78,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""OpenCompass or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: AdvancedIF = 79.78. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
claude-opus-4.8,advancedif,72.88,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""OpenCompass or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: AdvancedIF = 72.88. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
intern-s2-preview-397b,hmmt_feb_2026,91.57,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""max 256K text"",""effort"":""source does not state"",""harness"":""OpenCompass or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""thinking (default)"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",True,model_card,verified,Intern-S2-Preview-397B official model card: HMMT-2026 = 91.57. This is the release model's own reported evaluation.
kimi-k2.7-code,hmmt_feb_2026,90.34,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""OpenCompass or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: HMMT-2026 = 90.34. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
intern-s2-preview-397b,mmmu_pro,80.46,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""max 64K multimodal"",""effort"":""source does not state"",""harness"":""VLMEvalKit or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""thinking (default)"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",True,model_card,verified,Intern-S2-Preview-397B official model card: MMMU Pro = 80.46. This is the release model's own reported evaluation.
kimi-k2.7-code,mmmu_pro,77.92,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""VLMEvalKit or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: MMMU Pro = 77.92. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
intern-s2-preview-397b,chartqapro,69.65,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""max 64K multimodal"",""effort"":""source does not state"",""harness"":""VLMEvalKit or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""thinking (default)"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",True,model_card,verified,Intern-S2-Preview-397B official model card: ChartQAPro = 69.65. This is the release model's own reported evaluation.
qwen3.5-397b,chartqapro,68.61,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""VLMEvalKit or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: ChartQAPro = 68.61. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
kimi-k2.7-code,chartqapro,54.86,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""VLMEvalKit or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: ChartQAPro = 54.86. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
gpt-5.5,chartqapro,69.23,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""VLMEvalKit or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: ChartQAPro = 69.23. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
gemini-3.1-pro,chartqapro,71.18,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""VLMEvalKit or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: ChartQAPro = 71.18. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
claude-opus-4.8,chartqapro,58.65,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""VLMEvalKit or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: ChartQAPro = 58.65. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
intern-s2-preview-397b,terminal_bench_2_1,67.42,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""max 256K text"",""effort"":""source does not state"",""harness"":""Terminus 2"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""thinking (default)"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""terminal/code execution""}",True,model_card,verified,Intern-S2-Preview-397B official model card: TerminalBench 2.1 = 67.42. This is the release model's own reported evaluation.
qwen3.5-397b,terminal_bench_2_1,51.3,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""Terminus 2"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""terminal/code execution""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: TerminalBench 2.1 = 51.3. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
kimi-k2.7-code,terminal_bench_2_1,66.29,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""Terminus 2"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""terminal/code execution""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: TerminalBench 2.1 = 66.29. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
intern-s2-preview-397b,swe_bench_pro,61.56,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""max 256K text"",""effort"":""source does not state"",""harness"":""Mini-SWE-Agent"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""thinking (default)"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""agentic code execution""}",True,model_card,verified,Intern-S2-Preview-397B official model card: SWE-Bench-Pro = 61.56. This is the release model's own reported evaluation.
qwen3.5-397b,swe_bench_pro,43.55,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""Mini-SWE-Agent"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""agentic code execution""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: SWE-Bench-Pro = 43.55. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
kimi-k2.7-code,swe_bench_pro,57.59,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""Mini-SWE-Agent"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""agentic code execution""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: SWE-Bench-Pro = 57.59. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
intern-s2-preview-397b,swe_bench_multilingual,81.67,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""max 256K text"",""effort"":""source does not state"",""harness"":""Mini-SWE-Agent"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""thinking (default)"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""agentic code execution""}",True,model_card,verified,Intern-S2-Preview-397B official model card: SWE-Bench-Multilingual = 81.67. This is the release model's own reported evaluation.
kimi-k2.7-code,swe_bench_multilingual,78.56,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""Mini-SWE-Agent"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""agentic code execution""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: SWE-Bench-Multilingual = 78.56. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
intern-s2-preview-397b,xlrs_bench_micro,51.97,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""max 64K multimodal"",""effort"":""source does not state"",""harness"":""VLMEvalKit or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""thinking (default)"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",True,model_card,verified,Intern-S2-Preview-397B official model card: XLRS-Bench = 51.97. This is the release model's own reported evaluation.
qwen3.5-397b,xlrs_bench_micro,50.11,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""VLMEvalKit or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: XLRS-Bench = 50.11. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
kimi-k2.7-code,xlrs_bench_micro,49.9,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""VLMEvalKit or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: XLRS-Bench = 49.9. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
gpt-5.5,xlrs_bench_micro,50.96,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""VLMEvalKit or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: XLRS-Bench = 50.96. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
gemini-3.1-pro,xlrs_bench_micro,54.27,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""VLMEvalKit or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: XLRS-Bench = 54.27. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
claude-opus-4.8,xlrs_bench_micro,51.84,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""VLMEvalKit or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: XLRS-Bench = 51.84. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
intern-s2-preview-397b,microvqa,68.81,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""max 64K multimodal"",""effort"":""source does not state"",""harness"":""VLMEvalKit or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""thinking (default)"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",True,model_card,verified,Intern-S2-Preview-397B official model card: MicroVQA = 68.81. This is the release model's own reported evaluation.
qwen3.5-397b,microvqa,68.71,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""VLMEvalKit or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: MicroVQA = 68.71. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
kimi-k2.7-code,microvqa,61.04,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""VLMEvalKit or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: MicroVQA = 61.04. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
gpt-5.5,microvqa,63.63,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""VLMEvalKit or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: MicroVQA = 63.63. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
gemini-3.1-pro,microvqa,71.02,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""VLMEvalKit or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: MicroVQA = 71.02. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
claude-opus-4.8,microvqa,61.8,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""VLMEvalKit or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: MicroVQA = 61.8. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
intern-s2-preview-397b,sfe,61.67,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""max 64K multimodal"",""effort"":""source does not state"",""harness"":""VLMEvalKit or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""thinking (default)"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",True,model_card,verified,Intern-S2-Preview-397B official model card: SFE = 61.67. This is the release model's own reported evaluation.
qwen3.5-397b,sfe,62.97,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""VLMEvalKit or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: SFE = 62.97. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
kimi-k2.7-code,sfe,50.76,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""VLMEvalKit or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: SFE = 50.76. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
gpt-5.5,sfe,52.09,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""VLMEvalKit or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: SFE = 52.09. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
gemini-3.1-pro,sfe,59.57,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""VLMEvalKit or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: SFE = 59.57. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
claude-opus-4.8,sfe,59.08,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""VLMEvalKit or AgentCompass; exact adapter not stated"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: SFE = 59.08. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
intern-s2-preview-397b,scicode,49.11,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""max 256K text"",""effort"":""source does not state"",""harness"":""AgentCompass / benchmark evaluator"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""thinking (default)"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""code execution""}",True,model_card,verified,Intern-S2-Preview-397B official model card: SciCode = 49.11. This is the release model's own reported evaluation.
qwen3.5-397b,scicode,46.35,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""AgentCompass / benchmark evaluator"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""code execution""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: SciCode = 46.35. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
deepseek-v4-pro,scicode,47.53,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""AgentCompass / benchmark evaluator"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""code execution""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: SciCode = 47.53. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
kimi-k2.7-code,scicode,43.49,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""AgentCompass / benchmark evaluator"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""code execution""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: SciCode = 43.49. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
intern-s2-preview-397b,sgi_bench,49.37,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""max 64K multimodal"",""effort"":""source does not state"",""harness"":""official SGI-Bench agentic evaluation"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""thinking (default)"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""web search, PDF parser, Python, file reader""}",True,model_card,verified,Intern-S2-Preview-397B official model card: SGI-Bench = 49.37. This is the release model's own reported evaluation.
qwen3.5-397b,sgi_bench,44.44,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official SGI-Bench agentic evaluation"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""web search, PDF parser, Python, file reader""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: SGI-Bench = 44.44. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
deepseek-v4-pro,sgi_bench,45.7,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official SGI-Bench agentic evaluation"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""web search, PDF parser, Python, file reader""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: SGI-Bench = 45.7. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
kimi-k2.7-code,sgi_bench,50.63,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official SGI-Bench agentic evaluation"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""web search, PDF parser, Python, file reader""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: SGI-Bench = 50.63. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
glm-5.2,sgi_bench,52.41,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official SGI-Bench agentic evaluation"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""web search, PDF parser, Python, file reader""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: SGI-Bench = 52.41. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
gpt-5.5,sgi_bench,42.77,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official SGI-Bench agentic evaluation"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""web search, PDF parser, Python, file reader""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: SGI-Bench = 42.77. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
gemini-3.1-pro,sgi_bench,45.28,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official SGI-Bench agentic evaluation"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""web search, PDF parser, Python, file reader""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: SGI-Bench = 45.28. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
claude-opus-4.8,sgi_bench,49.06,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official SGI-Bench agentic evaluation"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""web search, PDF parser, Python, file reader""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: SGI-Bench = 49.06. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
intern-s2-preview-397b,researchclawbench,18.44,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""max 64K multimodal"",""effort"":""source does not state"",""harness"":""ResearchHarness"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""thinking (default)"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""scientific research environment""}",True,model_card,verified,Intern-S2-Preview-397B official model card: ResearchClawBench = 18.44. This is the release model's own reported evaluation.
qwen3.5-397b,researchclawbench,15.86,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""ResearchHarness"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""scientific research environment""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: ResearchClawBench = 15.86. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
deepseek-v4-pro,researchclawbench,13.69,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""ResearchHarness"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""scientific research environment""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: ResearchClawBench = 13.69. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
kimi-k2.7-code,researchclawbench,15.4,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""ResearchHarness"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""scientific research environment""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: ResearchClawBench = 15.4. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
glm-5.2,researchclawbench,23.35,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""ResearchHarness"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""scientific research environment""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: ResearchClawBench = 23.35. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
gpt-5.5,researchclawbench,17.0,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""ResearchHarness"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""scientific research environment""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: ResearchClawBench = 17.0. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
gemini-3.1-pro,researchclawbench,14.54,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""ResearchHarness"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""scientific research environment""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: ResearchClawBench = 14.54. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
claude-opus-4.8,researchclawbench,21.74,https://huggingface.co/internlm/Intern-S2-Preview-397B,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""ResearchHarness"",""judge"":""benchmark-specific; exact version not stated"",""mode"":""source does not state"",""notes"":""Shanghai AI Lab evaluated all model columns using OpenCompass, VLMEvalKit, and AgentCompass. Per-model mode/effort/sampling are not reported; only Intern-S2 maximum evaluation lengths are disclosed."",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""scientific research environment""}",False,model_card,verified,"Intern-S2-Preview-397B official model card: ResearchClawBench = 21.74. This is a Shanghai AI Lab cross-model evaluation; the source does not disclose this model's exact mode, effort, or sampling."
nemotron-3-ultra-550b-a55b,terminal_bench_2_1,56.4,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""Harbor / Terminus-2"",""sampling"":""8-run average"",""tools"":""terminal and code execution"",""version_metric"":""v2.1 / % resolved""}",True,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Nemotron 3 Ultra 550B-A55B: terminal_bench_2_1=56.4."
nemotron-3-ultra-550b-a55b,gdpval_normalized_elo,46.7,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / Stirrup"",""sampling"":""pass@1"",""tools"":""office workflow, web search, sandboxed code"",""version_metric"":""normalized Elo / 0-100""}",True,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Nemotron 3 Ultra 550B-A55B: gdpval_normalized_elo=46.7."
nemotron-3-ultra-550b-a55b,swe_bench_verified,70.7,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""Harbor / OpenHands / AWS ECS"",""sampling"":""3-run average"",""tools"":""agentic repository editing"",""version_metric"":""Verified / % resolved""}",True,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Nemotron 3 Ultra 550B-A55B: swe_bench_verified=70.7."
nemotron-3-ultra-550b-a55b,swe_bench_multilingual,67.7,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""Harbor / OpenHands / AWS ECS"",""sampling"":""3-run average"",""tools"":""agentic repository editing"",""version_metric"":""Multilingual / % resolved""}",True,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Nemotron 3 Ultra 550B-A55B: swe_bench_multilingual=67.7."
nemotron-3-ultra-550b-a55b,profbench,56.0,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""official/internal ProfBench scaffold"",""sampling"":""16-run average"",""tools"":""web search and browsing"",""version_metric"":""full / search score""}",True,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Nemotron 3 Ultra 550B-A55B: profbench=56."
nemotron-3-ultra-550b-a55b,pinchbench,90.0,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""official PinchBench scaffold"",""sampling"":""pass@1"",""tools"":""OpenClaw coding environment"",""version_metric"":""current 53-task release / %""}",True,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Nemotron 3 Ultra 550B-A55B: pinchbench=90."
nemotron-3-ultra-550b-a55b,tau3_airline,81.5,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / tau3"",""sampling"":""50 tasks x 8 trials"",""tools"":""domain customer-service tools"",""version_metric"":""airline / 8-trial average""}",True,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Nemotron 3 Ultra 550B-A55B: tau3_airline=81.5."
nemotron-3-ultra-550b-a55b,tau3_retail,86.4,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / tau3"",""sampling"":""114 tasks x 8 trials"",""tools"":""domain customer-service tools"",""version_metric"":""retail / 8-trial average""}",True,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Nemotron 3 Ultra 550B-A55B: tau3_retail=86.4."
nemotron-3-ultra-550b-a55b,tau3_telecom,92.9,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / tau3"",""sampling"":""114 tasks x 8 trials"",""tools"":""domain customer-service tools"",""version_metric"":""telecom / 8-trial average""}",True,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Nemotron 3 Ultra 550B-A55B: tau3_telecom=92.9."
nemotron-3-ultra-550b-a55b,tau3_banking,22.6,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / tau3"",""sampling"":""97 tasks x 8 trials"",""tools"":""domain tools plus terminal knowledge-base search"",""version_metric"":""banking_knowledge / 8-trial average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Nemotron 3 Ultra 550B-A55B: tau3_banking=22.6."
nemotron-3-ultra-550b-a55b,tau3_bench,70.9,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / tau3"",""sampling"":""375 tasks x 8 trials"",""tools"":""domain customer-service tools"",""version_metric"":""four-domain / 8-trial average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Nemotron 3 Ultra 550B-A55B: tau3_bench=70.9."
nemotron-3-ultra-550b-a55b,browsecomp,44.4,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NVIDIA custom BrowseComp scaffold"",""sampling"":""pass@1"",""tools"":""Tavily web search and terminal workspace"",""version_metric"":""BrowseComp / %""}",True,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Nemotron 3 Ultra 550B-A55B: browsecomp=44.4."
nemotron-3-ultra-550b-a55b,ioi_2025,570.0,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NVIDIA contest evaluation"",""sampling"":""pass@1"",""tools"":""code execution"",""version_metric"":""contest points / 600""}",True,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Nemotron 3 Ultra 550B-A55B: ioi_2025=570."
nemotron-3-ultra-550b-a55b,livecodebench_v6,89.0,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym"",""sampling"":""pass@1"",""tools"":""code execution"",""version_metric"":""v6 / pass@1""}",True,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Nemotron 3 Ultra 550B-A55B: livecodebench_v6=89."
nemotron-3-ultra-550b-a55b,imo_answerbench,88.6,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""5-run average"",""tools"":""none"",""variant"":""no tools"",""version_metric"":""no tools / %""}",True,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Nemotron 3 Ultra 550B-A55B: imo_answerbench=88.6, variant=no tools."
nemotron-3-ultra-550b-a55b,apex_shortlist,74.9,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""pass@1"",""tools"":""none"",""variant"":""no tools"",""version_metric"":""no tools / %""}",True,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Nemotron 3 Ultra 550B-A55B: apex_shortlist=74.9, variant=no tools."
nemotron-3-ultra-550b-a55b,gpqa_diamond,87.0,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""8-run average"",""tools"":""none"",""version_metric"":""no tools / %""}",True,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Nemotron 3 Ultra 550B-A55B: gpqa_diamond=87."
nemotron-3-ultra-550b-a55b,scicode,44.6,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""8-run average"",""tools"":""none"",""version_metric"":""subtask / %""}",True,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Nemotron 3 Ultra 550B-A55B: scicode=44.6."
nemotron-3-ultra-550b-a55b,hle_text,26.7,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""pass@1"",""tools"":""none"",""variant"":""no tools"",""version_metric"":""text-only / no tools / %""}",True,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Nemotron 3 Ultra 550B-A55B: hle_text=26.7, variant=no tools."
nemotron-3-ultra-550b-a55b,critpt,3.1,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""5-run average"",""tools"":""none"",""version_metric"":""no tools / %""}",True,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Nemotron 3 Ultra 550B-A55B: critpt=3.1."
nemotron-3-ultra-550b-a55b,mmlu_pro,86.8,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""MMLU-Pro / %""}",True,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Nemotron 3 Ultra 550B-A55B: mmlu_pro=86.8."
nemotron-3-ultra-550b-a55b,aa_omniscience_accuracy,24.1,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""10-run average"",""tools"":""none"",""version_metric"":""accuracy / %""}",True,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Nemotron 3 Ultra 550B-A55B: aa_omniscience_accuracy=24.1."
nemotron-3-ultra-550b-a55b,aa_omniscience_non_hallucination,78.7,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""10-run average"",""tools"":""none"",""version_metric"":""non-hallucination / %""}",True,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Nemotron 3 Ultra 550B-A55B: aa_omniscience_non_hallucination=78.7."
nemotron-3-ultra-550b-a55b,ifbench,81.7,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""8-run average"",""tools"":""none"",""version_metric"":""prompt-level loose accuracy / %""}",True,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Nemotron 3 Ultra 550B-A55B: ifbench=81.7."
nemotron-3-ultra-550b-a55b,multichallenge,63.8,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""8-run average"",""tools"":""none"",""version_metric"":""multi-turn score / %""}",True,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Nemotron 3 Ultra 550B-A55B: multichallenge=63.8."
nemotron-3-ultra-550b-a55b,aa_lcr,65.4,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""16-run average"",""tools"":""none"",""version_metric"":""AA-LCR / %""}",True,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Nemotron 3 Ultra 550B-A55B: aa_lcr=65.4."
nemotron-3-ultra-550b-a55b,ruler_1m,94.7,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Skills / RULER"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""1M / %""}",True,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Nemotron 3 Ultra 550B-A55B: ruler_1m=94.7."
nemotron-3-ultra-550b-a55b,longbench_v2,61.9,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""<=1M / %""}",True,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Nemotron 3 Ultra 550B-A55B: longbench_v2=61.9."
nemotron-3-ultra-550b-a55b,wmt24pp,83.7,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""54,890 translations"",""tools"":""none"",""version_metric"":""en-to-55 languages / XCOMET-XXL""}",True,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Nemotron 3 Ultra 550B-A55B: wmt24pp=83.7."
minimax-m2.7,terminal_bench_2_1,55.5,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""Harbor / Terminus-2"",""sampling"":""8-run average"",""tools"":""terminal and code execution"",""version_metric"":""v2.1 / % resolved""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: MiniMax M2.7: terminal_bench_2_1=55.5."
minimax-m2.7,gdpval_normalized_elo,47.6,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / Stirrup"",""sampling"":""pass@1"",""tools"":""office workflow, web search, sandboxed code"",""version_metric"":""normalized Elo / 0-100""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: MiniMax M2.7: gdpval_normalized_elo=47.6."
minimax-m2.7,swe_bench_verified,75.3,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""Harbor / OpenHands / AWS ECS"",""sampling"":""3-run average"",""tools"":""agentic repository editing"",""version_metric"":""Verified / % resolved""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: MiniMax M2.7: swe_bench_verified=75.3."
minimax-m2.7,profbench,52.0,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""official/internal ProfBench scaffold"",""sampling"":""16-run average"",""tools"":""web search and browsing"",""version_metric"":""full / search score""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: MiniMax M2.7: profbench=52."
minimax-m2.7,pinchbench,77.6,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""official PinchBench scaffold"",""sampling"":""pass@1"",""tools"":""OpenClaw coding environment"",""version_metric"":""current 53-task release / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: MiniMax M2.7: pinchbench=77.6."
minimax-m2.7,tau3_airline,75.3,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / tau3"",""sampling"":""50 tasks x 8 trials"",""tools"":""domain customer-service tools"",""version_metric"":""airline / 8-trial average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: MiniMax M2.7: tau3_airline=75.3."
minimax-m2.7,tau3_retail,84.9,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / tau3"",""sampling"":""114 tasks x 8 trials"",""tools"":""domain customer-service tools"",""version_metric"":""retail / 8-trial average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: MiniMax M2.7: tau3_retail=84.9."
minimax-m2.7,tau3_telecom,89.6,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / tau3"",""sampling"":""114 tasks x 8 trials"",""tools"":""domain customer-service tools"",""version_metric"":""telecom / 8-trial average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: MiniMax M2.7: tau3_telecom=89.6."
minimax-m2.7,tau3_banking,14.6,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / tau3"",""sampling"":""97 tasks x 8 trials"",""tools"":""domain tools plus terminal knowledge-base search"",""version_metric"":""banking_knowledge / 8-trial average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: MiniMax M2.7: tau3_banking=14.6."
minimax-m2.7,browsecomp,54.1,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NVIDIA custom BrowseComp scaffold"",""sampling"":""pass@1"",""tools"":""Tavily web search and terminal workspace"",""version_metric"":""BrowseComp / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: MiniMax M2.7: browsecomp=54.1."
minimax-m2.7,livecodebench_v6,77.2,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym"",""sampling"":""pass@1"",""tools"":""code execution"",""version_metric"":""v6 / pass@1""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: MiniMax M2.7: livecodebench_v6=77.2."
minimax-m2.7,apex_shortlist,28.9,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""pass@1"",""tools"":""none"",""variant"":""no tools"",""version_metric"":""no tools / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: MiniMax M2.7: apex_shortlist=28.9, variant=no tools."
minimax-m2.7,scicode,38.3,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""8-run average"",""tools"":""none"",""version_metric"":""subtask / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: MiniMax M2.7: scicode=38.3."
minimax-m2.7,hle_text,23.1,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""pass@1"",""tools"":""none"",""variant"":""no tools"",""version_metric"":""text-only / no tools / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: MiniMax M2.7: hle_text=23.1, variant=no tools."
minimax-m2.7,critpt,0.6,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""5-run average"",""tools"":""none"",""version_metric"":""no tools / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: MiniMax M2.7: critpt=0.6."
minimax-m2.7,mmlu_pro,81.9,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""MMLU-Pro / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: MiniMax M2.7: mmlu_pro=81.9."
minimax-m2.7,aa_omniscience_accuracy,20.5,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""10-run average"",""tools"":""none"",""version_metric"":""accuracy / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: MiniMax M2.7: aa_omniscience_accuracy=20.5."
minimax-m2.7,aa_omniscience_non_hallucination,74.4,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""10-run average"",""tools"":""none"",""version_metric"":""non-hallucination / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: MiniMax M2.7: aa_omniscience_non_hallucination=74.4."
minimax-m2.7,ifbench,74.6,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""8-run average"",""tools"":""none"",""version_metric"":""prompt-level loose accuracy / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: MiniMax M2.7: ifbench=74.6."
minimax-m2.7,multichallenge,42.5,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""8-run average"",""tools"":""none"",""version_metric"":""multi-turn score / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: MiniMax M2.7: multichallenge=42.5."
minimax-m2.7,aa_lcr,69.8,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""16-run average"",""tools"":""none"",""version_metric"":""AA-LCR / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: MiniMax M2.7: aa_lcr=69.8."
minimax-m2.7,wmt24pp,82.8,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""54,890 translations"",""tools"":""none"",""version_metric"":""en-to-55 languages / XCOMET-XXL""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: MiniMax M2.7: wmt24pp=82.8."
glm-5.1,gdpval_normalized_elo,54.7,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / Stirrup"",""sampling"":""pass@1"",""tools"":""office workflow, web search, sandboxed code"",""version_metric"":""normalized Elo / 0-100""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: GLM-5.1: gdpval_normalized_elo=54.7."
glm-5.1,profbench,46.0,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""official/internal ProfBench scaffold"",""sampling"":""16-run average"",""tools"":""web search and browsing"",""version_metric"":""full / search score""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: GLM-5.1: profbench=46."
glm-5.1,pinchbench,81.2,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""official PinchBench scaffold"",""sampling"":""pass@1"",""tools"":""OpenClaw coding environment"",""version_metric"":""current 53-task release / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: GLM-5.1: pinchbench=81.2."
glm-5.1,tau3_airline,85.0,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / tau3"",""sampling"":""50 tasks x 8 trials"",""tools"":""domain customer-service tools"",""version_metric"":""airline / 8-trial average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: GLM-5.1: tau3_airline=85."
glm-5.1,tau3_retail,84.1,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / tau3"",""sampling"":""114 tasks x 8 trials"",""tools"":""domain customer-service tools"",""version_metric"":""retail / 8-trial average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: GLM-5.1: tau3_retail=84.1."
glm-5.1,tau3_telecom,96.9,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / tau3"",""sampling"":""114 tasks x 8 trials"",""tools"":""domain customer-service tools"",""version_metric"":""telecom / 8-trial average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: GLM-5.1: tau3_telecom=96.9."
glm-5.1,tau3_banking,12.8,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / tau3"",""sampling"":""97 tasks x 8 trials"",""tools"":""domain tools plus terminal knowledge-base search"",""version_metric"":""banking_knowledge / 8-trial average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: GLM-5.1: tau3_banking=12.8."
glm-5.1,ioi_2025,456.5,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NVIDIA contest evaluation"",""sampling"":""pass@1"",""tools"":""code execution"",""version_metric"":""contest points / 600""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: GLM-5.1: ioi_2025=456.5."
glm-5.1,livecodebench_v6,85.7,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym"",""sampling"":""pass@1"",""tools"":""code execution"",""version_metric"":""v6 / pass@1""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: GLM-5.1: livecodebench_v6=85.7."
glm-5.1,scicode,47.7,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""8-run average"",""tools"":""none"",""version_metric"":""subtask / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: GLM-5.1: scicode=47.7."
glm-5.1,aa_omniscience_accuracy,31.3,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""10-run average"",""tools"":""none"",""version_metric"":""accuracy / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: GLM-5.1: aa_omniscience_accuracy=31.3."
glm-5.1,aa_omniscience_non_hallucination,66.8,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""10-run average"",""tools"":""none"",""version_metric"":""non-hallucination / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: GLM-5.1: aa_omniscience_non_hallucination=66.8."
glm-5.1,ifbench,76.6,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""8-run average"",""tools"":""none"",""version_metric"":""prompt-level loose accuracy / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: GLM-5.1: ifbench=76.6."
glm-5.1,multichallenge,63.0,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""8-run average"",""tools"":""none"",""version_metric"":""multi-turn score / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: GLM-5.1: multichallenge=63."
glm-5.1,wmt24pp,84.4,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""54,890 translations"",""tools"":""none"",""version_metric"":""en-to-55 languages / XCOMET-XXL""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: GLM-5.1: wmt24pp=84.4."
kimi-k2.6,terminal_bench_2_1,67.2,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""Harbor / Terminus-2"",""sampling"":""8-run average"",""tools"":""terminal and code execution"",""version_metric"":""v2.1 / % resolved""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Kimi K2.6: terminal_bench_2_1=67.2."
kimi-k2.6,gdpval_normalized_elo,50.4,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / Stirrup"",""sampling"":""pass@1"",""tools"":""office workflow, web search, sandboxed code"",""version_metric"":""normalized Elo / 0-100""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Kimi K2.6: gdpval_normalized_elo=50.4."
kimi-k2.6,profbench,56.0,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""official/internal ProfBench scaffold"",""sampling"":""16-run average"",""tools"":""web search and browsing"",""version_metric"":""full / search score""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Kimi K2.6: profbench=56."
kimi-k2.6,pinchbench,90.2,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""official PinchBench scaffold"",""sampling"":""pass@1"",""tools"":""OpenClaw coding environment"",""version_metric"":""current 53-task release / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Kimi K2.6: pinchbench=90.2."
kimi-k2.6,tau3_airline,85.8,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / tau3"",""sampling"":""50 tasks x 8 trials"",""tools"":""domain customer-service tools"",""version_metric"":""airline / 8-trial average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Kimi K2.6: tau3_airline=85.8."
kimi-k2.6,tau3_retail,82.9,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / tau3"",""sampling"":""114 tasks x 8 trials"",""tools"":""domain customer-service tools"",""version_metric"":""retail / 8-trial average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Kimi K2.6: tau3_retail=82.9."
kimi-k2.6,tau3_telecom,97.8,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / tau3"",""sampling"":""114 tasks x 8 trials"",""tools"":""domain customer-service tools"",""version_metric"":""telecom / 8-trial average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Kimi K2.6: tau3_telecom=97.8."
kimi-k2.6,tau3_banking,23.1,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / tau3"",""sampling"":""97 tasks x 8 trials"",""tools"":""domain tools plus terminal knowledge-base search"",""version_metric"":""banking_knowledge / 8-trial average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Kimi K2.6: tau3_banking=23.1."
kimi-k2.6,tau3_bench,72.4,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / tau3"",""sampling"":""375 tasks x 8 trials"",""tools"":""domain customer-service tools"",""version_metric"":""four-domain / 8-trial average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Kimi K2.6: tau3_bench=72.4."
kimi-k2.6,ioi_2025,585.0,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NVIDIA contest evaluation"",""sampling"":""pass@1"",""tools"":""code execution"",""version_metric"":""contest points / 600""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Kimi K2.6: ioi_2025=585."
kimi-k2.6,critpt,9.1,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""5-run average"",""tools"":""none"",""version_metric"":""no tools / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Kimi K2.6: critpt=9.1."
kimi-k2.6,aa_omniscience_accuracy,35.5,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""10-run average"",""tools"":""none"",""version_metric"":""accuracy / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Kimi K2.6: aa_omniscience_accuracy=35.5."
kimi-k2.6,aa_omniscience_non_hallucination,67.1,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""10-run average"",""tools"":""none"",""version_metric"":""non-hallucination / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Kimi K2.6: aa_omniscience_non_hallucination=67.1."
kimi-k2.6,ifbench,73.7,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""8-run average"",""tools"":""none"",""version_metric"":""prompt-level loose accuracy / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Kimi K2.6: ifbench=73.7."
kimi-k2.6,multichallenge,63.1,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""8-run average"",""tools"":""none"",""version_metric"":""multi-turn score / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Kimi K2.6: multichallenge=63.1."
kimi-k2.6,aa_lcr,70.2,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""16-run average"",""tools"":""none"",""version_metric"":""AA-LCR / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Kimi K2.6: aa_lcr=70.2."
kimi-k2.6,wmt24pp,84.5,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""54,890 translations"",""tools"":""none"",""version_metric"":""en-to-55 languages / XCOMET-XXL""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Kimi K2.6: wmt24pp=84.5."
qwen3.5-397b,gdpval_normalized_elo,34.6,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / Stirrup"",""sampling"":""pass@1"",""tools"":""office workflow, web search, sandboxed code"",""version_metric"":""normalized Elo / 0-100""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Qwen3.5-397B-A17B: gdpval_normalized_elo=34.6."
qwen3.5-397b,profbench,53.0,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""official/internal ProfBench scaffold"",""sampling"":""16-run average"",""tools"":""web search and browsing"",""version_metric"":""full / search score""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Qwen3.5-397B-A17B: profbench=53."
qwen3.5-397b,pinchbench,86.6,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""official PinchBench scaffold"",""sampling"":""pass@1"",""tools"":""OpenClaw coding environment"",""version_metric"":""current 53-task release / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Qwen3.5-397B-A17B: pinchbench=86.6."
qwen3.5-397b,tau3_airline,76.5,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / tau3"",""sampling"":""50 tasks x 8 trials"",""tools"":""domain customer-service tools"",""version_metric"":""airline / 8-trial average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Qwen3.5-397B-A17B: tau3_airline=76.5."
qwen3.5-397b,tau3_retail,88.5,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / tau3"",""sampling"":""114 tasks x 8 trials"",""tools"":""domain customer-service tools"",""version_metric"":""retail / 8-trial average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Qwen3.5-397B-A17B: tau3_retail=88.5."
qwen3.5-397b,tau3_telecom,98.0,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / tau3"",""sampling"":""114 tasks x 8 trials"",""tools"":""domain customer-service tools"",""version_metric"":""telecom / 8-trial average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Qwen3.5-397B-A17B: tau3_telecom=98."
qwen3.5-397b,tau3_banking,20.9,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / tau3"",""sampling"":""97 tasks x 8 trials"",""tools"":""domain tools plus terminal knowledge-base search"",""version_metric"":""banking_knowledge / 8-trial average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Qwen3.5-397B-A17B: tau3_banking=20.9."
qwen3.5-397b,tau3_bench,71.0,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / tau3"",""sampling"":""375 tasks x 8 trials"",""tools"":""domain customer-service tools"",""version_metric"":""four-domain / 8-trial average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Qwen3.5-397B-A17B: tau3_bench=71."
qwen3.5-397b,ioi_2025,441.3,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NVIDIA contest evaluation"",""sampling"":""pass@1"",""tools"":""code execution"",""version_metric"":""contest points / 600""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Qwen3.5-397B-A17B: ioi_2025=441.3."
qwen3.5-397b,livecodebench_v6,79.3,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym"",""sampling"":""pass@1"",""tools"":""code execution"",""version_metric"":""v6 / pass@1""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Qwen3.5-397B-A17B: livecodebench_v6=79.3."
qwen3.5-397b,apex_shortlist,61.4,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""pass@1"",""tools"":""none"",""variant"":""no tools"",""version_metric"":""no tools / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Qwen3.5-397B-A17B: apex_shortlist=61.4, variant=no tools."
qwen3.5-397b,hle_text,28.5,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""pass@1"",""tools"":""none"",""variant"":""no tools"",""version_metric"":""text-only / no tools / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Qwen3.5-397B-A17B: hle_text=28.5, variant=no tools."
qwen3.5-397b,critpt,2.4,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""5-run average"",""tools"":""none"",""version_metric"":""no tools / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Qwen3.5-397B-A17B: critpt=2.4."
qwen3.5-397b,aa_omniscience_accuracy,35.9,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""10-run average"",""tools"":""none"",""version_metric"":""accuracy / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Qwen3.5-397B-A17B: aa_omniscience_accuracy=35.9."
qwen3.5-397b,aa_omniscience_non_hallucination,7.4,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""10-run average"",""tools"":""none"",""version_metric"":""non-hallucination / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Qwen3.5-397B-A17B: aa_omniscience_non_hallucination=7.4."
qwen3.5-397b,ruler_1m,90.1,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Skills / RULER"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""1M / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Qwen3.5-397B-A17B: ruler_1m=90.1."
qwen3.5-397b,wmt24pp,86.8,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""54,890 translations"",""tools"":""none"",""version_metric"":""en-to-55 languages / XCOMET-XXL""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: Qwen3.5-397B-A17B: wmt24pp=86.8."
deepseek-v4-pro,gdpval_normalized_elo,54.6,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / Stirrup"",""sampling"":""pass@1"",""tools"":""office workflow, web search, sandboxed code"",""version_metric"":""normalized Elo / 0-100""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: DeepSeek-V4-Pro: gdpval_normalized_elo=54.6."
deepseek-v4-pro,profbench,59.9,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""official/internal ProfBench scaffold"",""sampling"":""16-run average"",""tools"":""web search and browsing"",""version_metric"":""full / search score""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: DeepSeek-V4-Pro: profbench=59.9."
deepseek-v4-pro,pinchbench,88.6,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""official PinchBench scaffold"",""sampling"":""pass@1"",""tools"":""OpenClaw coding environment"",""version_metric"":""current 53-task release / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: DeepSeek-V4-Pro: pinchbench=88.6."
deepseek-v4-pro,tau3_airline,80.8,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / tau3"",""sampling"":""50 tasks x 8 trials"",""tools"":""domain customer-service tools"",""version_metric"":""airline / 8-trial average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: DeepSeek-V4-Pro: tau3_airline=80.8."
deepseek-v4-pro,tau3_retail,88.9,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / tau3"",""sampling"":""114 tasks x 8 trials"",""tools"":""domain customer-service tools"",""version_metric"":""retail / 8-trial average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: DeepSeek-V4-Pro: tau3_retail=88.9."
deepseek-v4-pro,tau3_telecom,96.3,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / tau3"",""sampling"":""114 tasks x 8 trials"",""tools"":""domain customer-service tools"",""version_metric"":""telecom / 8-trial average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: DeepSeek-V4-Pro: tau3_telecom=96.3."
deepseek-v4-pro,tau3_banking,25.9,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / tau3"",""sampling"":""97 tasks x 8 trials"",""tools"":""domain tools plus terminal knowledge-base search"",""version_metric"":""banking_knowledge / 8-trial average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: DeepSeek-V4-Pro: tau3_banking=25.9."
deepseek-v4-pro,tau3_bench,73.2,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / tau3"",""sampling"":""375 tasks x 8 trials"",""tools"":""domain customer-service tools"",""version_metric"":""four-domain / 8-trial average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: DeepSeek-V4-Pro: tau3_bench=73.2."
deepseek-v4-pro,ioi_2025,580.1,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NVIDIA contest evaluation"",""sampling"":""pass@1"",""tools"":""code execution"",""version_metric"":""contest points / 600""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: DeepSeek-V4-Pro: ioi_2025=580.1."
deepseek-v4-pro,livecodebench_v6,93.5,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""default"",""effort"":""Max"",""harness"":""official"",""judge"":""rule-based"",""mode"":""thinking"",""notes"":""Per DeepSeek V4-Pro model card. Three reasoning modes available (Non-Think/High/Max); BP canonical = Max mode (most powerful). Other modes' values stored as candidates if needed."",""prompt_style"":""default"",""sampling"":""pass@1"",""temperature"":""default"",""tools"":""none (per-bench override)""}",True,official_paper,verified,Displayed exactly as 93.5. Official Microsoft-reported result. Comparator value is quoted from that provider's official model card/release; use the campaign canonical setting.
deepseek-v4-pro,aa_omniscience_accuracy,46.8,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""10-run average"",""tools"":""none"",""version_metric"":""accuracy / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: DeepSeek-V4-Pro: aa_omniscience_accuracy=46.8."
deepseek-v4-pro,aa_omniscience_non_hallucination,5.7,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""10-run average"",""tools"":""none"",""version_metric"":""non-hallucination / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: DeepSeek-V4-Pro: aa_omniscience_non_hallucination=5.7."
deepseek-v4-pro,ifbench,79.1,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""8-run average"",""tools"":""none"",""version_metric"":""prompt-level loose accuracy / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: DeepSeek-V4-Pro: ifbench=79.1."
deepseek-v4-pro,multichallenge,64.1,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""8-run average"",""tools"":""none"",""version_metric"":""multi-turn score / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: DeepSeek-V4-Pro: multichallenge=64.1."
deepseek-v4-pro,ruler_1m,94.2,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Skills / RULER"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""1M / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: DeepSeek-V4-Pro: ruler_1m=94.2."
deepseek-v4-pro,longbench_v2,62.1,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""<=1M / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: DeepSeek-V4-Pro: longbench_v2=62.1."
deepseek-v4-pro,wmt24pp,85.9,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""54,890 translations"",""tools"":""none"",""version_metric"":""en-to-55 languages / XCOMET-XXL""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: DeepSeek-V4-Pro: wmt24pp=85.9."
deepseek-v4-flash,gdpval_normalized_elo,50.2,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / Stirrup"",""sampling"":""pass@1"",""tools"":""office workflow, web search, sandboxed code"",""version_metric"":""normalized Elo / 0-100""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: DeepSeek-V4-Flash Preview: gdpval_normalized_elo=50.2."
deepseek-v4-flash,profbench,57.0,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""official/internal ProfBench scaffold"",""sampling"":""16-run average"",""tools"":""web search and browsing"",""version_metric"":""full / search score""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: DeepSeek-V4-Flash Preview: profbench=57."
deepseek-v4-flash,pinchbench,91.3,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""official PinchBench scaffold"",""sampling"":""pass@1"",""tools"":""OpenClaw coding environment"",""version_metric"":""current 53-task release / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: DeepSeek-V4-Flash Preview: pinchbench=91.3."
deepseek-v4-flash,tau3_airline,80.8,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / tau3"",""sampling"":""50 tasks x 8 trials"",""tools"":""domain customer-service tools"",""version_metric"":""airline / 8-trial average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: DeepSeek-V4-Flash Preview: tau3_airline=80.8."
deepseek-v4-flash,tau3_retail,89.1,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / tau3"",""sampling"":""114 tasks x 8 trials"",""tools"":""domain customer-service tools"",""version_metric"":""retail / 8-trial average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: DeepSeek-V4-Flash Preview: tau3_retail=89.1."
deepseek-v4-flash,tau3_telecom,98.3,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / tau3"",""sampling"":""114 tasks x 8 trials"",""tools"":""domain customer-service tools"",""version_metric"":""telecom / 8-trial average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: DeepSeek-V4-Flash Preview: tau3_telecom=98.3."
deepseek-v4-flash,tau3_banking,26.7,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / tau3"",""sampling"":""97 tasks x 8 trials"",""tools"":""domain tools plus terminal knowledge-base search"",""version_metric"":""banking_knowledge / 8-trial average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: DeepSeek-V4-Flash Preview: tau3_banking=26.7."
deepseek-v4-flash,tau3_bench,73.7,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym / tau3"",""sampling"":""375 tasks x 8 trials"",""tools"":""domain customer-service tools"",""version_metric"":""four-domain / 8-trial average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: DeepSeek-V4-Flash Preview: tau3_bench=73.7."
deepseek-v4-flash,livecodebench_v6,90.9,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Gym"",""sampling"":""pass@1"",""tools"":""code execution"",""version_metric"":""v6 / pass@1""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: DeepSeek-V4-Flash Preview: livecodebench_v6=90.9."
deepseek-v4-flash,scicode,48.2,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""8-run average"",""tools"":""none"",""version_metric"":""subtask / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: DeepSeek-V4-Flash Preview: scicode=48.2."
deepseek-v4-flash,critpt,10.6,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""5-run average"",""tools"":""none"",""version_metric"":""no tools / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: DeepSeek-V4-Flash Preview: critpt=10.6."
deepseek-v4-flash,aa_omniscience_accuracy,39.9,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""10-run average"",""tools"":""none"",""version_metric"":""accuracy / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: DeepSeek-V4-Flash Preview: aa_omniscience_accuracy=39.9."
deepseek-v4-flash,aa_omniscience_non_hallucination,2.8,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""10-run average"",""tools"":""none"",""version_metric"":""non-hallucination / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: DeepSeek-V4-Flash Preview: aa_omniscience_non_hallucination=2.8."
deepseek-v4-flash,ifbench,82.0,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""8-run average"",""tools"":""none"",""version_metric"":""prompt-level loose accuracy / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: DeepSeek-V4-Flash Preview: ifbench=82."
deepseek-v4-flash,multichallenge,63.5,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""8-run average"",""tools"":""none"",""version_metric"":""multi-turn score / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: DeepSeek-V4-Flash Preview: multichallenge=63.5."
deepseek-v4-flash,ruler_1m,87.7,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Skills / RULER"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""1M / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: DeepSeek-V4-Flash Preview: ruler_1m=87.7."
deepseek-v4-flash,longbench_v2,57.0,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""<=1M / %""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: DeepSeek-V4-Flash Preview: longbench_v2=57."
deepseek-v4-flash,wmt24pp,85.9,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NeMo Evaluator"",""sampling"":""54,890 translations"",""tools"":""none"",""version_metric"":""en-to-55 languages / XCOMET-XXL""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 10: DeepSeek-V4-Flash Preview: wmt24pp=85.9."
nemotron-3-ultra-550b-a55b,mmlu,89.08,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""5-shot accuracy""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Nemotron 3 Ultra Base: mmlu=89.08. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
nemotron-3-ultra-550b-a55b,agieval_en,78.73,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""3/5-shot CoT exact match""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Nemotron 3 Ultra Base: agieval_en=78.73. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
nemotron-3-ultra-550b-a55b,gsm8k,88.1,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""8-shot CoT exact match""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Nemotron 3 Ultra Base: gsm8k=88.1. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
nemotron-3-ultra-550b-a55b,math_test,82.0,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""Minerva 4-shot exact match""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Nemotron 3 Ultra Base: math_test=82. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
nemotron-3-ultra-550b-a55b,humaneval,83.84,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""164 problems x 32 samples"",""tools"":""none"",""version_metric"":""EvalPlus sanitized sampled pass@1 n=32""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Nemotron 3 Ultra Base: humaneval=83.84. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
nemotron-3-ultra-550b-a55b,mbpp_sanitized,85.97,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""427 problems x 32 samples"",""tools"":""none"",""version_metric"":""3-shot sampled pass@1 n=32""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Nemotron 3 Ultra Base: mbpp_sanitized=85.97. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
nemotron-3-ultra-550b-a55b,arc_challenge,97.35,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""25-shot normalized accuracy""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Nemotron 3 Ultra Base: arc_challenge=97.35. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
nemotron-3-ultra-550b-a55b,hellaswag,90.51,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""10-shot normalized accuracy""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Nemotron 3 Ultra Base: hellaswag=90.51. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
nemotron-3-ultra-550b-a55b,openbookqa,48.6,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""0-shot normalized accuracy""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Nemotron 3 Ultra Base: openbookqa=48.6. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
nemotron-3-ultra-550b-a55b,piqa,83.79,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""0-shot normalized accuracy""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Nemotron 3 Ultra Base: piqa=83.79. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
nemotron-3-ultra-550b-a55b,winogrande,79.32,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""5-shot accuracy""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Nemotron 3 Ultra Base: winogrande=79.32. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
nemotron-3-ultra-550b-a55b,race,92.15,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""0-shot accuracy""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Nemotron 3 Ultra Base: race=92.15. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
nemotron-3-ultra-550b-a55b,global_mmlu_lite,90.13,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""5-shot average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Nemotron 3 Ultra Base: global_mmlu_lite=90.13. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
nemotron-3-ultra-550b-a55b,mgsm,87.73,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""8-shot native CoT average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Nemotron 3 Ultra Base: mgsm=87.73. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
nemotron-3-ultra-550b-a55b,ruler_64k,95.3,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Skills / RULER"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""0-shot""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Nemotron 3 Ultra Base: ruler_64k=95.3. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
nemotron-3-ultra-550b-a55b,ruler_128k,92.49,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Skills / RULER"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""0-shot""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Nemotron 3 Ultra Base: ruler_128k=92.49. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
nemotron-3-ultra-550b-a55b,ruler_256k,86.22,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Skills / RULER"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""0-shot""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Nemotron 3 Ultra Base: ruler_256k=86.22. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
nemotron-3-ultra-550b-a55b,ruler_512k,84.54,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Skills / RULER"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""0-shot""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Nemotron 3 Ultra Base: ruler_512k=84.54. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
deepseek-v3.2,agieval_en,70.13,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""3/5-shot CoT exact match""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: DeepSeek-V3.2-Exp-Base: agieval_en=70.13. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
deepseek-v3.2,gsm8k,84.38,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""8-shot CoT exact match""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: DeepSeek-V3.2-Exp-Base: gsm8k=84.38. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
deepseek-v3.2,math_test,60.12,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""Minerva 4-shot exact match""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: DeepSeek-V3.2-Exp-Base: math_test=60.12. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
deepseek-v3.2,mbpp_sanitized,58.66,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""427 problems x 32 samples"",""tools"":""none"",""version_metric"":""3-shot sampled pass@1 n=32""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: DeepSeek-V3.2-Exp-Base: mbpp_sanitized=58.66. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
deepseek-v3.2,arc_challenge,95.22,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""25-shot normalized accuracy""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: DeepSeek-V3.2-Exp-Base: arc_challenge=95.22. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
deepseek-v3.2,hellaswag,89.44,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""10-shot normalized accuracy""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: DeepSeek-V3.2-Exp-Base: hellaswag=89.44. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
deepseek-v3.2,openbookqa,48.2,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""0-shot normalized accuracy""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: DeepSeek-V3.2-Exp-Base: openbookqa=48.2. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
deepseek-v3.2,piqa,85.09,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""0-shot normalized accuracy""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: DeepSeek-V3.2-Exp-Base: piqa=85.09. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
deepseek-v3.2,winogrande,83.43,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""5-shot accuracy""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: DeepSeek-V3.2-Exp-Base: winogrande=83.43. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
deepseek-v3.2,race,93.21,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""0-shot accuracy""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: DeepSeek-V3.2-Exp-Base: race=93.21. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
deepseek-v3.2,global_mmlu_lite,85.59,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""5-shot average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: DeepSeek-V3.2-Exp-Base: global_mmlu_lite=85.59. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
deepseek-v3.2,mgsm,82.33,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""8-shot native CoT average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: DeepSeek-V3.2-Exp-Base: mgsm=82.33. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
deepseek-v3.2,ruler_64k,93.3,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Skills / RULER"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""0-shot""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: DeepSeek-V3.2-Exp-Base: ruler_64k=93.3. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
deepseek-v3.2,ruler_128k,91.88,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Skills / RULER"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""0-shot""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: DeepSeek-V3.2-Exp-Base: ruler_128k=91.88. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
mistral-large-3,agieval_en,69.3,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""3/5-shot CoT exact match""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Mistral Large 3 675B Base 2512: agieval_en=69.3. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
mistral-large-3,gsm8k,91.21,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""8-shot CoT exact match""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Mistral Large 3 675B Base 2512: gsm8k=91.21. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
mistral-large-3,math_test,62.88,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""Minerva 4-shot exact match""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Mistral Large 3 675B Base 2512: math_test=62.88. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
mistral-large-3,mbpp_sanitized,84.08,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""427 problems x 32 samples"",""tools"":""none"",""version_metric"":""3-shot sampled pass@1 n=32""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Mistral Large 3 675B Base 2512: mbpp_sanitized=84.08. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
mistral-large-3,arc_challenge,97.27,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""25-shot normalized accuracy""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Mistral Large 3 675B Base 2512: arc_challenge=97.27. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
mistral-large-3,hellaswag,88.88,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""10-shot normalized accuracy""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Mistral Large 3 675B Base 2512: hellaswag=88.88. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
mistral-large-3,openbookqa,51.4,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""0-shot normalized accuracy""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Mistral Large 3 675B Base 2512: openbookqa=51.4. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
mistral-large-3,piqa,84.82,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""0-shot normalized accuracy""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Mistral Large 3 675B Base 2512: piqa=84.82. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
mistral-large-3,winogrande,82.08,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""5-shot accuracy""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Mistral Large 3 675B Base 2512: winogrande=82.08. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
mistral-large-3,race,93.3,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""0-shot accuracy""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Mistral Large 3 675B Base 2512: race=93.3. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
mistral-large-3,global_mmlu_lite,87.34,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""5-shot average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Mistral Large 3 675B Base 2512: global_mmlu_lite=87.34. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
mistral-large-3,mgsm,82.93,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""8-shot native CoT average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Mistral Large 3 675B Base 2512: mgsm=82.93. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
mistral-large-3,ruler_64k,90.11,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Skills / RULER"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""0-shot""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Mistral Large 3 675B Base 2512: ruler_64k=90.11. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
mistral-large-3,ruler_128k,55.77,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Skills / RULER"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""0-shot""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Mistral Large 3 675B Base 2512: ruler_128k=55.77. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
mistral-large-3,ruler_256k,35.5,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Skills / RULER"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""0-shot""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Mistral Large 3 675B Base 2512: ruler_256k=35.5. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
kimi-k2,agieval_en,72.55,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""3/5-shot CoT exact match""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Kimi K2 Base: agieval_en=72.55. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
kimi-k2,math_test,68.4,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""Minerva 4-shot exact match""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Kimi K2 Base: math_test=68.4. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
kimi-k2,mbpp_sanitized,72.14,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""427 problems x 32 samples"",""tools"":""none"",""version_metric"":""3-shot sampled pass@1 n=32""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Kimi K2 Base: mbpp_sanitized=72.14. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
kimi-k2,arc_challenge,95.82,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""25-shot normalized accuracy""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Kimi K2 Base: arc_challenge=95.82. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
kimi-k2,hellaswag,90.92,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""10-shot normalized accuracy""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Kimi K2 Base: hellaswag=90.92. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
kimi-k2,openbookqa,50.8,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""0-shot normalized accuracy""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Kimi K2 Base: openbookqa=50.8. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
kimi-k2,piqa,85.47,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""0-shot normalized accuracy""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Kimi K2 Base: piqa=85.47. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
kimi-k2,winogrande,84.21,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""5-shot accuracy""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Kimi K2 Base: winogrande=84.21. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
kimi-k2,race,91.96,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""0-shot accuracy""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Kimi K2 Base: race=91.96. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
kimi-k2,global_mmlu_lite,85.63,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""5-shot average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Kimi K2 Base: global_mmlu_lite=85.63. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
kimi-k2,mgsm,85.2,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""8-shot native CoT average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Kimi K2 Base: mgsm=85.2. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
kimi-k2,ruler_64k,93.79,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Skills / RULER"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""0-shot""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Kimi K2 Base: ruler_64k=93.79. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
kimi-k2,ruler_128k,88.61,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Skills / RULER"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""0-shot""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: Kimi K2 Base: ruler_128k=88.61. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
glm-4.5,mmlu,86.5,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""5-shot accuracy""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: GLM-4.5 Base: mmlu=86.5. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
glm-4.5,mmlu_pro,65.78,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""5-shot CoT exact match""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: GLM-4.5 Base: mmlu_pro=65.78. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
glm-4.5,agieval_en,70.06,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""3/5-shot CoT exact match""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: GLM-4.5 Base: agieval_en=70.06. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
glm-4.5,gpqa_diamond,34.85,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""5-shot CoT exact match""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: GLM-4.5 Base: gpqa_diamond=34.85. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
glm-4.5,gsm8k,85.37,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""8-shot CoT exact match""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: GLM-4.5 Base: gsm8k=85.37. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
glm-4.5,math_test,57.58,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""Minerva 4-shot exact match""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: GLM-4.5 Base: math_test=57.58. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
glm-4.5,humaneval,78.16,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""164 problems x 32 samples"",""tools"":""none"",""version_metric"":""EvalPlus sanitized sampled pass@1 n=32""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: GLM-4.5 Base: humaneval=78.16. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
glm-4.5,mbpp_sanitized,76.69,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""427 problems x 32 samples"",""tools"":""none"",""version_metric"":""3-shot sampled pass@1 n=32""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: GLM-4.5 Base: mbpp_sanitized=76.69. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
glm-4.5,arc_challenge,96.59,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""25-shot normalized accuracy""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: GLM-4.5 Base: arc_challenge=96.59. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
glm-4.5,hellaswag,90.17,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""10-shot normalized accuracy""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: GLM-4.5 Base: hellaswag=90.17. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
glm-4.5,openbookqa,49.6,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""0-shot normalized accuracy""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: GLM-4.5 Base: openbookqa=49.6. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
glm-4.5,piqa,85.09,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""0-shot normalized accuracy""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: GLM-4.5 Base: piqa=85.09. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
glm-4.5,winogrande,85.24,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""5-shot accuracy""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: GLM-4.5 Base: winogrande=85.24. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
glm-4.5,race,92.15,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""0-shot accuracy""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: GLM-4.5 Base: race=92.15. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
glm-4.5,global_mmlu_lite,85.81,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""5-shot average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: GLM-4.5 Base: global_mmlu_lite=85.81. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
glm-4.5,mgsm,81.27,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Evaluator"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""8-shot native CoT average""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: GLM-4.5 Base: mgsm=81.27. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
glm-4.5,ruler_64k,16.12,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Skills / RULER"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""0-shot""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: GLM-4.5 Base: ruler_64k=16.12. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
glm-4.5,ruler_128k,0.0,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""checkpoint"":""base"",""harness"":""NeMo Skills / RULER"",""mode"":""pretrained"",""sampling"":""pass@1"",""tools"":""none"",""version_metric"":""0-shot""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Base Table 2: GLM-4.5 Base: ruler_128k=0. Base checkpoint score mapped to the canonical post-trained model row as checkpoint=base."
nemotron-3-ultra-550b-a55b,imo_proofbench_advanced,82.3,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NVIDIA test-time-scaling math pipeline"",""sampling"":""five refinement rounds"",""tools"":""search and proof-refinement pipeline"",""variant"":""R5 final"",""version_metric"":""final five-round percentage""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 11: Nemotron 3 Ultra: imo_proofbench_advanced=82.3, variant=R5 final."
nemotron-3-ultra-550b-a55b,imo_2025,83.3,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NVIDIA test-time-scaling math pipeline"",""sampling"":""five refinement rounds"",""tools"":""search and proof-refinement pipeline"",""variant"":""R5 final"",""version_metric"":""final five-round percentage""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 11: Nemotron 3 Ultra: imo_2025=83.3, variant=R5 final."
nemotron-3-ultra-550b-a55b,putnam_2025,96.7,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NVIDIA test-time-scaling math pipeline"",""sampling"":""five refinement rounds"",""tools"":""search and proof-refinement pipeline"",""variant"":""R5 final"",""version_metric"":""final five-round percentage""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 11: Nemotron 3 Ultra: putnam_2025=96.7, variant=R5 final."
nemotron-3-ultra-550b-a55b,usamo_2026,97.6,https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3-Ultra-Technical-Report.pdf,"{""harness"":""NVIDIA test-time-scaling math pipeline"",""sampling"":""five refinement rounds"",""tools"":""search and proof-refinement pipeline"",""variant"":""R5 final"",""version_metric"":""final five-round percentage""}",False,tech_report,verified,"NVIDIA Nemotron 3 Ultra technical report, Final Table 11: Nemotron 3 Ultra: usamo_2026=97.6, variant=R5 final."
gpt-5.6-terra,agents_last_exam,50.4,https://openai.com/index/gpt-5-6/,"{""code_mode"":""true"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 0): GPT‑5.6 Terra; Agents' Last Exam=50.4. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-luna,agents_last_exam,50.3,https://openai.com/index/gpt-5-6/,"{""code_mode"":""true"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 0): GPT‑5.6 Luna; Agents' Last Exam=50.3. Exact effort/variant resolved from the rendered chart point.
gemini-3.1-pro,agents_last_exam,32.1,https://openai.com/index/gpt-5-6/,"{""code_mode"":""true"",""effort"":""xhigh"",""eval_variant"":null}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 0): Gemini 3.1 Pro Preview; Agents' Last Exam=32.1. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-terra,gdpval_aa_elo,1593.0,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 0): GPT‑5.6 Terra; GDPval-AA v2=1593.0. OpenAI row explicitly identifies GDPval-AA v2. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-luna,gdpval_aa_elo,1591.8,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 0): GPT‑5.6 Luna; GDPval-AA v2=1591.8. OpenAI row explicitly identifies GDPval-AA v2. Exact effort/variant resolved from the rendered chart point.
gemini-3.5-flash,gdpval_aa_elo,1348.8,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""high"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 0): Gemini 3.5 Flash; GDPval-AA v2=1348.8. OpenAI row explicitly identifies GDPval-AA v2. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-sol,management_consulting_tasks_internal,43.2,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 0): GPT‑5.6 Sol; Management Consulting Tasks (Internal)=43.2. Named internal evaluation; task count not disclosed.
gpt-5.6-terra,management_consulting_tasks_internal,37.2,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 0): GPT‑5.6 Terra; Management Consulting Tasks (Internal)=37.2. Named internal evaluation; task count not disclosed.
gpt-5.6-luna,management_consulting_tasks_internal,35.4,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 0): GPT‑5.6 Luna; Management Consulting Tasks (Internal)=35.4. Named internal evaluation; task count not disclosed.
gpt-5.5,management_consulting_tasks_internal,31.3,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 0): GPT‑5.5; Management Consulting Tasks (Internal)=31.3. Named internal evaluation; task count not disclosed.
claude-fable-5,management_consulting_tasks_internal,35.5,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 0): Claude Fable 5; Management Consulting Tasks (Internal)=35.5. Named internal evaluation; task count not disclosed.
claude-opus-4.8,management_consulting_tasks_internal,31.6,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 0): Claude Opus 4.8; Management Consulting Tasks (Internal)=31.6. Named internal evaluation; task count not disclosed.
gemini-3.1-pro,management_consulting_tasks_internal,13.2,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 0): Gemini 3.1 Pro Preview; Management Consulting Tasks (Internal)=13.2. Named internal evaluation; task count not disclosed.
gpt-5.6-sol,big_finance_bench,53.0,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 0): GPT‑5.6 Sol; Big Finance Bench=53.0. Distinct from the existing Finance Bench row.
gpt-5.6-terra,big_finance_bench,51.0,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 0): GPT‑5.6 Terra; Big Finance Bench=51.0. Distinct from the existing Finance Bench row.
gpt-5.6-luna,big_finance_bench,36.0,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 0): GPT‑5.6 Luna; Big Finance Bench=36.0. Distinct from the existing Finance Bench row.
gpt-5.5,big_finance_bench,49.0,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 0): GPT‑5.5; Big Finance Bench=49.0. Distinct from the existing Finance Bench row.
claude-opus-4.8,big_finance_bench,44.0,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 0): Claude Opus 4.8; Big Finance Bench=44.0. Distinct from the existing Finance Bench row.
gpt-5.6-terra,aa_intelligence_index_v4_1,55.0,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 0): GPT‑5.6 Terra; Artificial Analysis Intelligence Index v4.1=55.0. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-luna,aa_intelligence_index_v4_1,51.2,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 0): GPT‑5.6 Luna; Artificial Analysis Intelligence Index v4.1=51.2. Exact effort/variant resolved from the rendered chart point.
gemini-3.1-pro,aa_intelligence_index_v4_1,46.5,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""n/a"",""eval_variant"":null}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 0): Gemini 3.1 Pro Preview; Artificial Analysis Intelligence Index v4.1=46.5. Exact effort/variant resolved from the rendered chart point.
gemini-3.5-flash,aa_intelligence_index_v4_1,50.2,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""high"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 0): Gemini 3.5 Flash; Artificial Analysis Intelligence Index v4.1=50.2. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-sol,aa_coding_agent_index_v1_1,80.0,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 1): GPT‑5.6 Sol; Artificial Analysis Coding Agent Index v1.1=80.0. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-terra,aa_coding_agent_index_v1_1,77.4,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 1): GPT‑5.6 Terra; Artificial Analysis Coding Agent Index v1.1=77.4. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-luna,aa_coding_agent_index_v1_1,74.6,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 1): GPT‑5.6 Luna; Artificial Analysis Coding Agent Index v1.1=74.6. Exact effort/variant resolved from the rendered chart point.
gpt-5.5,aa_coding_agent_index_v1_1,76.4,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""xhigh"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 1): GPT‑5.5; Artificial Analysis Coding Agent Index v1.1=76.4. Exact effort/variant resolved from the rendered chart point.
claude-fable-5,aa_coding_agent_index_v1_1,77.2,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 1): Claude Fable 5; Artificial Analysis Coding Agent Index v1.1=77.2. Exact effort/variant resolved from the rendered chart point.
claude-opus-4.8,aa_coding_agent_index_v1_1,72.5,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 1): Claude Opus 4.8; Artificial Analysis Coding Agent Index v1.1=72.5. Exact effort/variant resolved from the rendered chart point.
gemini-3.1-pro,aa_coding_agent_index_v1_1,42.7,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""n/a"",""eval_variant"":null}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 1): Gemini 3.1 Pro Preview; Artificial Analysis Coding Agent Index v1.1=42.7. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-sol,swe_bench_pro,64.6,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 1): GPT‑5.6 Sol; SWE-Bench Pro=64.6.
gpt-5.6-terra,swe_bench_pro,63.4,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 1): GPT‑5.6 Terra; SWE-Bench Pro=63.4.
gpt-5.6-luna,swe_bench_pro,62.7,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 1): GPT‑5.6 Luna; SWE-Bench Pro=62.7.
claude-fable-5,swe_bench_pro,80.4,https://media.x.ai/v1/website/4p5-5184fdf9.pdf,"{""context"":""unknown"",""effort"":""max, with fallback"",""harness"":""fixed SWE-bench Pro agent scaffold"",""judge"":""verification reward and repository tests"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic repository shell/editor""}",True,model_card,verified,"card45 SWE-Bench Pro Resolve rate (%); source effort=max, with fallback; benchmark protocol matches canonical=True. Unknown source fields remain unknown."
gpt-5.6-terra,deep_swe_v1_1,69.6,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 1): GPT‑5.6 Terra; DeepSWE v1.1=69.6. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-luna,deep_swe_v1_1,67.2,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 1): GPT‑5.6 Luna; DeepSWE v1.1=67.2. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-terra,terminal_bench_2_1,87.4,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 1): GPT‑5.6 Terra; Terminal-Bench 2.1=87.4.
gpt-5.6-luna,terminal_bench_2_1,84.7,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 1): GPT‑5.6 Luna; Terminal-Bench 2.1=84.7.
gpt-5.6-sol,genebench_pro,28.7,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 2): GPT‑5.6 Sol; GeneBench Pro=28.7. Distinct professional variant from the existing GeneBench row. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-terra,genebench_pro,23.3,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 2): GPT‑5.6 Terra; GeneBench Pro=23.3. Distinct professional variant from the existing GeneBench row. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-luna,genebench_pro,10.8,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 2): GPT‑5.6 Luna; GeneBench Pro=10.8. Distinct professional variant from the existing GeneBench row. Exact effort/variant resolved from the rendered chart point.
gpt-5.5,genebench_pro,12.0,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""xhigh"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 2): GPT‑5.5; GeneBench Pro=12.0. Distinct professional variant from the existing GeneBench row. Exact effort/variant resolved from the rendered chart point.
claude-opus-4.8,genebench_pro,16.0,https://openai.com/index/gpt-5-6/,"{""code_mode"":null,""effort"":""reference"",""eval_variant"":null}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 2): Claude Opus 4.8; GeneBench Pro=16.0. Distinct professional variant from the existing GeneBench row. Exact effort/variant resolved from the rendered chart point.
gemini-3.1-pro,genebench_pro,3.1,https://openai.com/index/gpt-5-6/,"{""code_mode"":null,""effort"":""reference"",""eval_variant"":null}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 2): Gemini 3.1 Pro Preview; GeneBench Pro=3.1. Distinct professional variant from the existing GeneBench row. Exact effort/variant resolved from the rendered chart point.
gemini-3.5-flash,genebench_pro,8.14,https://openai.com/index/gpt-5-6/,"{""code_mode"":null,""effort"":""reference"",""eval_variant"":null}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 2): Gemini 3.5 Flash; GeneBench Pro=8.14. Distinct professional variant from the existing GeneBench row. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-sol,lifescibench,59.9,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 2): GPT‑5.6 Sol; LifeSciBench=59.9. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-terra,lifescibench,56.0,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 2): GPT‑5.6 Terra; LifeSciBench=56.0. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-luna,lifescibench,51.2,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 2): GPT‑5.6 Luna; LifeSciBench=51.2. Exact effort/variant resolved from the rendered chart point.
gpt-5.5,lifescibench,50.4,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""xhigh"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 2): GPT‑5.5; LifeSciBench=50.4. Exact effort/variant resolved from the rendered chart point.
claude-opus-4.8,lifescibench,53.6,https://openai.com/index/gpt-5-6/,"{""code_mode"":null,""effort"":""reference"",""eval_variant"":null}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 2): Claude Opus 4.8; LifeSciBench=53.6. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-sol,medchembench_internal,48.3,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 2): GPT‑5.6 Sol; MedChemBench (Internal)=48.3. Named internal evaluation; task count not disclosed. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-terra,medchembench_internal,35.0,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 2): GPT‑5.6 Terra; MedChemBench (Internal)=35.0. Named internal evaluation; task count not disclosed. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-luna,medchembench_internal,30.4,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 2): GPT‑5.6 Luna; MedChemBench (Internal)=30.4. Named internal evaluation; task count not disclosed. Exact effort/variant resolved from the rendered chart point.
gpt-5.5,medchembench_internal,35.5,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""xhigh"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 2): GPT‑5.5; MedChemBench (Internal)=35.5. Named internal evaluation; task count not disclosed. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-sol,healthbench_professional_length_adjusted,60.5,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 2): GPT‑5.6 Sol; HealthBench Professional⁶=60.5. OpenAI's official length-adjusted scoring; not comparable to Anthropic system-card HealthBench Professional scores.
gpt-5.6-terra,healthbench_professional_length_adjusted,57.7,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 2): GPT‑5.6 Terra; HealthBench Professional⁶=57.7. OpenAI's official length-adjusted scoring; not comparable to Anthropic system-card HealthBench Professional scores.
gpt-5.6-luna,healthbench_professional_length_adjusted,55.7,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 2): GPT‑5.6 Luna; HealthBench Professional⁶=55.7. OpenAI's official length-adjusted scoring; not comparable to Anthropic system-card HealthBench Professional scores.
gpt-5.5,healthbench_professional_length_adjusted,49.5,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 2): GPT‑5.5; HealthBench Professional⁶=49.5. OpenAI's official length-adjusted scoring; not comparable to Anthropic system-card HealthBench Professional scores.
claude-fable-5,healthbench_professional_length_adjusted,60.9,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 2): Claude Fable 5; HealthBench Professional⁶=60.9. OpenAI's official length-adjusted scoring; not comparable to Anthropic system-card HealthBench Professional scores.
claude-opus-4.8,healthbench_professional_length_adjusted,53.0,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 2): Claude Opus 4.8; HealthBench Professional⁶=53.0. OpenAI's official length-adjusted scoring; not comparable to Anthropic system-card HealthBench Professional scores.
gpt-5.6-terra,osworld_2_0,50.2,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 3): GPT‑5.6 Terra; OSWorld 2.0=50.2. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-luna,osworld_2_0,45.6,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 3): GPT‑5.6 Luna; OSWorld 2.0=45.6. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-terra,browsecomp,87.5,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 3): GPT‑5.6 Terra; BrowseComp=87.5. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-luna,browsecomp,83.3,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 3): GPT‑5.6 Luna; BrowseComp=83.3. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-sol,benchcad,70.6,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 3): GPT‑5.6 Sol; BenchCAD=70.6.
gpt-5.6-terra,benchcad,62.3,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 3): GPT‑5.6 Terra; BenchCAD=62.3.
gpt-5.6-luna,benchcad,63.1,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 3): GPT‑5.6 Luna; BenchCAD=63.1.
gpt-5.5,benchcad,44.4,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 3): GPT‑5.5; BenchCAD=44.4.
claude-fable-5,benchcad,38.4,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""deployment"":""claude-mythos-5"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Anthropic documents Mythos 5 as the same underlying model and capabilities as Fable 5, without Fable's safety classifiers."",""prompt_style"":""source does not state"",""safeguards"":""lifted for Project Glasswing"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 3): Claude Mythos 5; BenchCAD=38.4.
claude-mythos,benchcad,35.5,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 3): Claude Mythos Preview; BenchCAD=35.5.
claude-opus-4.8,benchcad,27.3,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 3): Claude Opus 4.8; BenchCAD=27.3.
gpt-5.6-sol,ctf_internal,96.67,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 4): GPT‑5.6 Sol; Capture-the-Flag Challenges=96.67. System Card identifies the release-page row as Capture the Flag (Internal). Exact effort/variant resolved from the rendered chart point. Stored at the System Card figure's exact printed precision; release table displays 96.7%.
gpt-5.6-terra,ctf_internal,91.84,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 4): GPT‑5.6 Terra; Capture-the-Flag Challenges=91.84. System Card identifies the release-page row as Capture the Flag (Internal). Exact effort/variant resolved from the rendered chart point. Stored at the System Card figure's exact printed precision; release table displays 91.8%.
gpt-5.6-luna,ctf_internal,85.19,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 4): GPT‑5.6 Luna; Capture-the-Flag Challenges=85.19. System Card identifies the release-page row as Capture the Flag (Internal). Exact effort/variant resolved from the rendered chart point. Stored at the System Card figure's exact printed precision; release table displays 85.2%.
gpt-5.6-sol,sec_bench_pro,71.2,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 4): GPT‑5.6 Sol; SEC-Bench Pro=71.2. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-terra,sec_bench_pro,57.7,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 4): GPT‑5.6 Terra; SEC-Bench Pro=57.7. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-luna,sec_bench_pro,48.9,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 4): GPT‑5.6 Luna; SEC-Bench Pro=48.9. Exact effort/variant resolved from the rendered chart point.
gpt-5.5,sec_bench_pro,45.8,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""xhigh"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 4): GPT‑5.5; SEC-Bench Pro=45.8. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-sol,exploitbench,73.5,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 4): GPT‑5.6 Sol; ExploitBench=73.5. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-terra,exploitbench,52.9,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 4): GPT‑5.6 Terra; ExploitBench=52.9. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-luna,exploitbench,33.2,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 4): GPT‑5.6 Luna; ExploitBench=33.2. Exact effort/variant resolved from the rendered chart point.
gpt-5.5,exploitbench,47.9,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""xhigh"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 4): GPT‑5.5; ExploitBench=47.9. Exact effort/variant resolved from the rendered chart point.
claude-fable-5,exploitbench,78.0,https://openai.com/index/gpt-5-6/,"{""code_mode"":null,""deployment"":""claude-mythos-5"",""effort"":""API, 5 seeds"",""eval_variant"":null,""safeguards"":""lifted for Project Glasswing""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 4): Claude Mythos 5; ExploitBench=78.0. Exact effort/variant resolved from the rendered chart point.
claude-mythos,exploitbench,74.2,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""n/a"",""eval_variant"":null}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 4): Claude Mythos Preview; ExploitBench=74.2. Exact effort/variant resolved from the rendered chart point.
claude-opus-4.8,exploitbench,40.0,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""exploit"",""sampling"":""pass@1"",""tools"":""cyber execution environment""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 3.2.1.A, page 31: Claude Opus 4.8; ExploitBench [Cap%]=40. Source setting: effort=source does not state; tools=cyber execution environment; sampling=pass@1; harness=exploit."
gpt-5.6-sol,exploitgym,33.7,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":""6h""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 4): GPT‑5.6 Sol; ExploitGym=33.7. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-terra,exploitgym,23.2,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":""6h""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 4): GPT‑5.6 Terra; ExploitGym=23.2. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-luna,exploitgym,12.4,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":""6h""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 4): GPT‑5.6 Luna; ExploitGym=12.4. Exact effort/variant resolved from the rendered chart point.
gpt-5.5,exploitgym,15.1,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""xhigh"",""eval_variant"":""2h""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 4): GPT‑5.5; ExploitGym=15.1. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-sol,internal_research_debugging,68.3,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 5): GPT‑5.6 Sol; Internal Research Debugging Evaluation=68.3. Named internal evaluation; task count not disclosed. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-terra,internal_research_debugging,67.8,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 5): GPT‑5.6 Terra; Internal Research Debugging Evaluation=67.8. Named internal evaluation; task count not disclosed. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-luna,internal_research_debugging,50.8,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 5): GPT‑5.6 Luna; Internal Research Debugging Evaluation=50.8. Named internal evaluation; task count not disclosed. Exact effort/variant resolved from the rendered chart point.
gpt-5.5,internal_research_debugging,50.0,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""xhigh"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 5): GPT‑5.5; Internal Research Debugging Evaluation=50.0. Named internal evaluation; task count not disclosed. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-sol,kernelgen_1p,61.1,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 5): GPT‑5.6 Sol; KernelGen 1P=61.1. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-terra,kernelgen_1p,49.2,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 5): GPT‑5.6 Terra; KernelGen 1P=49.2. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-luna,kernelgen_1p,22.4,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 5): GPT‑5.6 Luna; KernelGen 1P=22.4. Exact effort/variant resolved from the rendered chart point.
gpt-5.5,kernelgen_1p,29.3,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""xhigh"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 5): GPT‑5.5; KernelGen 1P=29.3. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-sol,nanogpt_self_improvement,9.69,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 5): GPT‑5.6 Sol; NanoGPT=9.69. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-terra,nanogpt_self_improvement,14.5,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 5): GPT‑5.6 Terra; NanoGPT=14.5. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-luna,nanogpt_self_improvement,1.66,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 5): GPT‑5.6 Luna; NanoGPT=1.66. Exact effort/variant resolved from the rendered chart point.
gpt-5.5,nanogpt_self_improvement,2.65,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""xhigh"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 5): GPT‑5.5; NanoGPT=2.65. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-sol,posttrain_bench_lite,50.3,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 5): GPT‑5.6 Sol; PostTrainBench Lite=50.3. Distinct Lite variant from the existing PostTrainBench row.
gpt-5.6-terra,posttrain_bench_lite,51.5,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 5): GPT‑5.6 Terra; PostTrainBench Lite=51.5. Distinct Lite variant from the existing PostTrainBench row.
gpt-5.6-luna,posttrain_bench_lite,29.6,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 5): GPT‑5.6 Luna; PostTrainBench Lite=29.6. Distinct Lite variant from the existing PostTrainBench row.
gpt-5.5,posttrain_bench_lite,38.8,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 5): GPT‑5.5; PostTrainBench Lite=38.8. Distinct Lite variant from the existing PostTrainBench row.
gpt-5.6-sol,rsi_index,57.9,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 5): GPT‑5.6 Sol; RSI Index=57.9. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-terra,rsi_index,56.3,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 5): GPT‑5.6 Terra; RSI Index=56.3. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-luna,rsi_index,41.9,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 5): GPT‑5.6 Luna; RSI Index=41.9. Exact effort/variant resolved from the rendered chart point.
gpt-5.5,rsi_index,41.7,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""xhigh"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 5): GPT‑5.5; RSI Index=41.7. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-terra,mmmu_pro,80.7,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 6): GPT‑5.6 Terra; MMMU Pro (no tools)=80.7.
gpt-5.6-luna,mmmu_pro,78.4,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 6): GPT‑5.6 Luna; MMMU Pro (no tools)=78.4.
gpt-5.6-sol,gdp_pdf,30.7,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 6): GPT‑5.6 Sol; gdp.pdf=30.7.
gpt-5.6-terra,gdp_pdf,24.7,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 6): GPT‑5.6 Terra; gdp.pdf=24.7.
gpt-5.6-luna,gdp_pdf,22.7,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 6): GPT‑5.6 Luna; gdp.pdf=22.7.
gpt-5.5,gdp_pdf,26.0,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 6): GPT‑5.5; gdp.pdf=26.0.
claude-fable-5,gdp_pdf,29.8,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 6): Claude Fable 5; gdp.pdf=29.8.
claude-opus-4.8,gdp_pdf,22.5,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 6): Claude Opus 4.8; gdp.pdf=22.5.
gemini-3.1-pro,gdp_pdf,16.7,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 6): Gemini 3.1 Pro Preview; gdp.pdf=16.7.
gpt-5.6-terra,gpqa_diamond,92.9,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 7): GPT‑5.6 Terra; GPQA Diamond=92.9.
gpt-5.6-luna,gpqa_diamond,92.3,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 7): GPT‑5.6 Luna; GPQA Diamond=92.3.
gpt-5.6-sol,frontiermath_tier_1_3_v2,89.0,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 7): GPT‑5.6 Sol; FrontierMath Tier 1-3 (v2)=89.0. Version 2 is distinct from the existing FrontierMath row.
gpt-5.6-terra,frontiermath_tier_1_3_v2,84.9,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 7): GPT‑5.6 Terra; FrontierMath Tier 1-3 (v2)=84.9. Version 2 is distinct from the existing FrontierMath row.
gpt-5.6-luna,frontiermath_tier_1_3_v2,78.6,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 7): GPT‑5.6 Luna; FrontierMath Tier 1-3 (v2)=78.6. Version 2 is distinct from the existing FrontierMath row.
gpt-5.5,frontiermath_tier_1_3_v2,85.3,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 7): GPT‑5.5; FrontierMath Tier 1-3 (v2)=85.3. Version 2 is distinct from the existing FrontierMath row.
claude-fable-5,frontiermath_tier_1_3_v2,87.0,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 7): Claude Fable 5; FrontierMath Tier 1-3 (v2)=87.0. Version 2 is distinct from the existing FrontierMath row.
claude-opus-4.8,frontiermath_tier_1_3_v2,80.0,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 7): Claude Opus 4.8; FrontierMath Tier 1-3 (v2)=80.0. Version 2 is distinct from the existing FrontierMath row.
gemini-3.1-pro,frontiermath_tier_1_3_v2,59.6,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 7): Gemini 3.1 Pro Preview; FrontierMath Tier 1-3 (v2)=59.6. Version 2 is distinct from the existing FrontierMath row.
gpt-5.6-sol,frontiermath_tier_4_v2,83.0,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 7): GPT‑5.6 Sol; FrontierMath Tier 4 (v2)=83.0. Version 2 is distinct from the existing FrontierMath Tier 4 row.
gpt-5.6-terra,frontiermath_tier_4_v2,68.3,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 7): GPT‑5.6 Terra; FrontierMath Tier 4 (v2)=68.3. Version 2 is distinct from the existing FrontierMath Tier 4 row.
gpt-5.6-luna,frontiermath_tier_4_v2,58.5,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 7): GPT‑5.6 Luna; FrontierMath Tier 4 (v2)=58.5. Version 2 is distinct from the existing FrontierMath Tier 4 row.
gpt-5.5,frontiermath_tier_4_v2,72.5,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 7): GPT‑5.5; FrontierMath Tier 4 (v2)=72.5. Version 2 is distinct from the existing FrontierMath Tier 4 row.
claude-fable-5,frontiermath_tier_4_v2,87.8,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 7): Claude Fable 5; FrontierMath Tier 4 (v2)=87.8. Version 2 is distinct from the existing FrontierMath Tier 4 row.
claude-opus-4.8,frontiermath_tier_4_v2,56.1,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 7): Claude Opus 4.8; FrontierMath Tier 4 (v2)=56.1. Version 2 is distinct from the existing FrontierMath Tier 4 row.
gpt-5.6-terra,automation_bench,15.2,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 8): GPT‑5.6 Terra; AutomationBench=15.2. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-luna,automation_bench,14.9,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""max"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 8): GPT‑5.6 Luna; AutomationBench=14.9. Exact effort/variant resolved from the rendered chart point.
gemini-3.5-flash,automation_bench,14.5,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""high"",""eval_variant"":null}",True,official_blog,verified,OpenAI GPT-5.6 release table (table 8): Gemini 3.5 Flash; AutomationBench=14.5. Exact effort/variant resolved from the rendered chart point.
gpt-5.6-sol,toolathlon,58.0,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 8): GPT‑5.6 Sol; Toolathlon=58.0.
gpt-5.6-terra,toolathlon,53.1,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 8): GPT‑5.6 Terra; Toolathlon=53.1.
gpt-5.6-luna,toolathlon,53.4,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 8): GPT‑5.6 Luna; Toolathlon=53.4.
claude-mythos,toolathlon,61.1,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""toolathlon"",""sampling"":""3 trials/task"",""tools"":""agentic benchmark harness""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 8.11.5.A, page 135: Claude Mythos Preview; Toolathlon [Pass@1]=61.1. Source setting: effort=max; tools=agentic benchmark harness; sampling=3 trials/task; harness=toolathlon."
claude-fable-5,toolathlon,61.7,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""toolathlon"",""sampling"":""3 trials/task"",""tools"":""agentic benchmark harness""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 8.11.5.A, page 135: Claude Fable 5; Toolathlon [Pass@1]=61.7. Source setting: effort=max; tools=agentic benchmark harness; sampling=3 trials/task; harness=toolathlon."
gpt-5.6-sol,mrcr_v2_8needle_256k_512k,91.5,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 9): GPT‑5.6 Sol; OpenAI MRCR v2 8-needle 256K-512K=91.5.
gpt-5.6-terra,mrcr_v2_8needle_256k_512k,89.6,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 9): GPT‑5.6 Terra; OpenAI MRCR v2 8-needle 256K-512K=89.6.
gpt-5.6-luna,mrcr_v2_8needle_256k_512k,41.3,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 9): GPT‑5.6 Luna; OpenAI MRCR v2 8-needle 256K-512K=41.3.
gpt-5.5,mrcr_v2_8needle_256k_512k,81.5,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 9): GPT‑5.5; OpenAI MRCR v2 8-needle 256K-512K=81.5.
gpt-5.6-sol,mrcr_v2_8needle,73.8,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,"OpenAI GPT-5.6 release table (table 9): GPT‑5.6 Sol; OpenAI MRCR v2 8-needle 512K-1M=73.8. The existing generic 8-needle row already contains GPT-5.5=74.0, which exactly matches this 512K-1M source row."
gpt-5.6-terra,mrcr_v2_8needle,72.5,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,"OpenAI GPT-5.6 release table (table 9): GPT‑5.6 Terra; OpenAI MRCR v2 8-needle 512K-1M=72.5. The existing generic 8-needle row already contains GPT-5.5=74.0, which exactly matches this 512K-1M source row."
gpt-5.6-luna,mrcr_v2_8needle,41.3,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,"OpenAI GPT-5.6 release table (table 9): GPT‑5.6 Luna; OpenAI MRCR v2 8-needle 512K-1M=41.3. The existing generic 8-needle row already contains GPT-5.5=74.0, which exactly matches this 512K-1M source row."
gpt-5.6-sol,graphwalks_bfs_256k,90.7,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 9): GPT‑5.6 Sol; GraphWalks BFS 256k f1=90.7. Distinct point from the existing 256K-1M averaged row.
gpt-5.6-terra,graphwalks_bfs_256k,76.9,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 9): GPT‑5.6 Terra; GraphWalks BFS 256k f1=76.9. Distinct point from the existing 256K-1M averaged row.
gpt-5.6-luna,graphwalks_bfs_256k,81.3,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 9): GPT‑5.6 Luna; GraphWalks BFS 256k f1=81.3. Distinct point from the existing 256K-1M averaged row.
gpt-5.5,graphwalks_bfs_256k,73.7,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 9): GPT‑5.5; GraphWalks BFS 256k f1=73.7. Distinct point from the existing 256K-1M averaged row.
claude-fable-5,graphwalks_bfs_256k,91.1,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""deployment"":""claude-mythos-5"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Anthropic documents Mythos 5 as the same underlying model and capabilities as Fable 5, without Fable's safety classifiers."",""prompt_style"":""source does not state"",""safeguards"":""lifted for Project Glasswing"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 9): Claude Mythos 5; GraphWalks BFS 256k f1=91.1. Distinct point from the existing 256K-1M averaged row.
claude-mythos,graphwalks_bfs_256k,85.7,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 9): Claude Mythos Preview; GraphWalks BFS 256k f1=85.7. Distinct point from the existing 256K-1M averaged row.
claude-opus-4.8,graphwalks_bfs_256k,85.9,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 9): Claude Opus 4.8; GraphWalks BFS 256k f1=85.9. Distinct point from the existing 256K-1M averaged row.
gpt-5.6-sol,graphwalks_bfs_1m,77.1,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 9): GPT‑5.6 Sol; GraphWalks BFS 1mil f1=77.1. Distinct point from the existing 256K-1M averaged row.
gpt-5.6-terra,graphwalks_bfs_1m,71.2,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 9): GPT‑5.6 Terra; GraphWalks BFS 1mil f1=71.2. Distinct point from the existing 256K-1M averaged row.
gpt-5.6-luna,graphwalks_bfs_1m,51.2,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 9): GPT‑5.6 Luna; GraphWalks BFS 1mil f1=51.2. Distinct point from the existing 256K-1M averaged row.
gpt-5.5,graphwalks_bfs_1m,45.4,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 9): GPT‑5.5; GraphWalks BFS 1mil f1=45.4. Distinct point from the existing 256K-1M averaged row.
claude-fable-5,graphwalks_bfs_1m,79.4,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""deployment"":""claude-mythos-5"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""notes"":""Anthropic documents Mythos 5 as the same underlying model and capabilities as Fable 5, without Fable's safety classifiers."",""prompt_style"":""source does not state"",""safeguards"":""lifted for Project Glasswing"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 9): Claude Mythos 5; GraphWalks BFS 1mil f1=79.4. Distinct point from the existing 256K-1M averaged row.
claude-mythos,graphwalks_bfs_1m,74.3,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 9): Claude Mythos Preview; GraphWalks BFS 1mil f1=74.3. Distinct point from the existing 256K-1M averaged row.
claude-opus-4.8,graphwalks_bfs_1m,68.1,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 9): Claude Opus 4.8; GraphWalks BFS 1mil f1=68.1. Distinct point from the existing 256K-1M averaged row.
gpt-5.6-sol,arc_agi_3,7.78,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 10): GPT‑5.6 Sol; ARC-AGI-3⁷=7.78.
gpt-5.6-terra,arc_agi_3,0.8,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 10): GPT‑5.6 Terra; ARC-AGI-3⁷=0.8.
gpt-5.6-luna,arc_agi_3,0.18,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 10): GPT‑5.6 Luna; ARC-AGI-3⁷=0.18.
gpt-5.5,arc_agi_3,0.43,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 10): GPT‑5.5; ARC-AGI-3⁷=0.43.
claude-opus-4.8,arc_agi_3,1.52,https://arcprize.org/leaderboard,"{""effort"":""high""}",True,leaderboard,verified,ARC Prize leaderboard / Claude Opus 4.8 (High): Relative Human Action Efficiency; total cost $10000.
gemini-3.1-pro,arc_agi_3,0.42,https://openai.com/index/gpt-5-6/,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official release evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state"",""temperature"":""source does not state"",""tools"":""none""}",False,official_blog,verified,OpenAI GPT-5.6 release table (table 10): Gemini 3.1 Pro Preview; ARC-AGI-3⁷=0.42.
gpt-5.4,automation_bench,7.3,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""xhigh"",""eval_variant"":null}",True,official_blog,verified,"OpenAI GPT-5.6 release figure (AutomationBench): GPT-5.4; AutomationBench=7.3. Figure 15: AutomationBench. Cost, latency, and output-token coordinates are retained as review metadata, not score-matrix metrics."
gpt-5.4,exploitbench,38.0,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""xhigh"",""eval_variant"":null}",True,official_blog,verified,"OpenAI GPT-5.6 release figure (ExploitBench): GPT-5.4; ExploitBench=38.0. Figure 27: Progress from vulnerable code to arbitrary code execution.. Cost, latency, and output-token coordinates are retained as review metadata, not score-matrix metrics."
gpt-5.4,exploitgym,7.0,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""xhigh"",""eval_variant"":""2h""}",False,official_blog,verified,"OpenAI GPT-5.6 release figure (ExploitGym): GPT-5.4 · 2h cap; ExploitGym=7.0. Figure 27: ExploitGym. Cost, latency, and output-token coordinates are retained as review metadata, not score-matrix metrics."
gpt-5.4,sec_bench_pro,35.93,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""xhigh"",""eval_variant"":null}",True,official_blog,verified,"OpenAI GPT-5.6 release figure (SEC-Bench Pro): GPT-5.4; SEC-Bench Pro=35.93. Figure 27: SEC-Bench Pro. Cost, latency, and output-token coordinates are retained as review metadata, not score-matrix metrics."
gpt-5.4,genebench_pro,8.87,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""xhigh"",""eval_variant"":null}",True,official_blog,verified,"OpenAI GPT-5.6 release figure (GeneBench Pro): GPT-5.4; GeneBench Pro=8.87. Figure 28: Long-horizon genomics and quantitative-biology analyses.. Cost, latency, and output-token coordinates are retained as review metadata, not score-matrix metrics."
gpt-5.4,lifescibench,44.17,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""xhigh"",""eval_variant"":null}",True,official_blog,verified,"OpenAI GPT-5.6 release figure (LifeSciBench): GPT-5.4; LifeSciBench=44.17. Figure 28: LifeSciBench. Cost, latency, and output-token coordinates are retained as review metadata, not score-matrix metrics."
gpt-5.4,medchembench_internal,23.98,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""xhigh"",""eval_variant"":null}",True,official_blog,verified,"OpenAI GPT-5.6 release figure (MedChemBench): GPT-5.4; MedChemBench (Internal)=23.98. Figure 28: MedChemBench. Cost, latency, and output-token coordinates are retained as review metadata, not score-matrix metrics."
gpt-5.4,rsi_index,36.36,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""xhigh"",""eval_variant"":null}",True,official_blog,verified,"OpenAI GPT-5.6 release figure (RSI Index): GPT-5.4; RSI Index=36.36. Figure 29: Aggregate capability on evaluations related to recursive self-improvement.. Cost, latency, and output-token coordinates are retained as review metadata, not score-matrix metrics."
gpt-5.4,internal_research_debugging,47.71,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""xhigh"",""eval_variant"":null}",True,official_blog,verified,"OpenAI GPT-5.6 release figure (Internal Research Debugging Evaluation): GPT-5.4; Internal Research Debugging Evaluation=47.71. Figure 29: Internal Research Debugging Evaluation. Cost, latency, and output-token coordinates are retained as review metadata, not score-matrix metrics."
gpt-5.4,kernelgen_1p,17.96,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""xhigh"",""eval_variant"":null}",True,official_blog,verified,"OpenAI GPT-5.6 release figure (KernelGen 1P): GPT-5.4; KernelGen 1P=17.96. Figure 29: KernelGen 1P. Cost, latency, and output-token coordinates are retained as review metadata, not score-matrix metrics."
gpt-5.4,nanogpt_self_improvement,2.3,https://openai.com/index/gpt-5-6/,"{""code_mode"":""false"",""effort"":""xhigh"",""eval_variant"":null}",True,official_blog,verified,"OpenAI GPT-5.6 release figure (NanoGPT): GPT-5.4; NanoGPT=2.3. Figure 29: NanoGPT. Cost, latency, and output-token coordinates are retained as review metadata, not score-matrix metrics."
gpt-5,healthbench_professional_length_adjusted,46.2,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5; HealthBench Professional length-adjusted=46.2. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5,healthbench_professional,51.0,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5; HealthBench Professional unadjusted=51.0. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.1,healthbench_professional_length_adjusted,39.6,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.1; HealthBench Professional length-adjusted=39.6. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.1,healthbench_professional,48.0,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.1; HealthBench Professional unadjusted=48.0. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.2,healthbench_professional_length_adjusted,45.9,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.2; HealthBench Professional length-adjusted=45.9. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.2,healthbench_professional,50.0,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.2; HealthBench Professional unadjusted=50.0. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.4,healthbench_professional_length_adjusted,48.1,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.4; HealthBench Professional length-adjusted=48.1. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.4,healthbench_professional,51.9,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.4; HealthBench Professional unadjusted=51.9. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.5,healthbench_professional,57.2,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.5; HealthBench Professional unadjusted=57.2. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.6-sol,healthbench_professional,64.1,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.6-sol; HealthBench Professional unadjusted=64.1. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.6-terra,healthbench_professional,62.4,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.6-terra; HealthBench Professional unadjusted=62.4. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.6-luna,healthbench_professional,59.8,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.6-luna; HealthBench Professional unadjusted=59.8. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5,healthbench_length_adjusted,57.7,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5; HealthBench length-adjusted=57.7. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.1,healthbench_length_adjusted,50.9,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.1; HealthBench length-adjusted=50.9. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.1,healthbench,64.2,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.1; HealthBench unadjusted=64.2. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.2,healthbench_length_adjusted,56.8,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.2; HealthBench length-adjusted=56.8. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.4,healthbench_length_adjusted,54.0,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.4; HealthBench length-adjusted=54.0. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.4,healthbench,55.7,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.4; HealthBench unadjusted=55.7. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.5,healthbench_length_adjusted,56.5,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.5; HealthBench length-adjusted=56.5. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.5,healthbench,58.4,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.5; HealthBench unadjusted=58.4. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.6-sol,healthbench_length_adjusted,57.0,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.6-sol; HealthBench length-adjusted=57.0. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.6-sol,healthbench,55.6,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.6-sol; HealthBench unadjusted=55.6. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.6-terra,healthbench_length_adjusted,57.0,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.6-terra; HealthBench length-adjusted=57.0. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.6-terra,healthbench,58.7,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.6-terra; HealthBench unadjusted=58.7. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.6-luna,healthbench_length_adjusted,55.8,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.6-luna; HealthBench length-adjusted=55.8. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.6-luna,healthbench,55.4,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.6-luna; HealthBench unadjusted=55.4. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5,healthbench_hard_length_adjusted,34.7,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5; HealthBench Hard length-adjusted=34.7. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.1,healthbench_hard_length_adjusted,25.4,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.1; HealthBench Hard length-adjusted=25.4. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.1,healthbench_hard,41.4,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.1; HealthBench Hard unadjusted=41.4. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.2,healthbench_hard_length_adjusted,34.3,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.2; HealthBench Hard length-adjusted=34.3. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.4,healthbench_hard_length_adjusted,29.1,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.4; HealthBench Hard length-adjusted=29.1. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.4,healthbench_hard,30.3,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.4; HealthBench Hard unadjusted=30.3. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.5,healthbench_hard_length_adjusted,31.5,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.5; HealthBench Hard length-adjusted=31.5. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.5,healthbench_hard,33.8,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.5; HealthBench Hard unadjusted=33.8. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.6-sol,healthbench_hard_length_adjusted,33.1,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.6-sol; HealthBench Hard length-adjusted=33.1. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.6-sol,healthbench_hard,31.1,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.6-sol; HealthBench Hard unadjusted=31.1. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.6-terra,healthbench_hard_length_adjusted,32.7,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.6-terra; HealthBench Hard length-adjusted=32.7. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.6-terra,healthbench_hard,34.3,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.6-terra; HealthBench Hard unadjusted=34.3. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.6-luna,healthbench_hard_length_adjusted,32.0,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.6-luna; HealthBench Hard length-adjusted=32.0. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.6-luna,healthbench_hard,31.4,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.6-luna; HealthBench Hard unadjusted=31.4. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5,healthbench_consensus_length_adjusted,95.6,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5; HealthBench Consensus length-adjusted=95.6. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5,healthbench_consensus,96.0,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5; HealthBench Consensus unadjusted=96.0. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.1,healthbench_consensus_length_adjusted,95.0,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.1; HealthBench Consensus length-adjusted=95.0. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.1,healthbench_consensus,95.8,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.1; HealthBench Consensus unadjusted=95.8. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.2,healthbench_consensus_length_adjusted,94.4,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.2; HealthBench Consensus length-adjusted=94.4. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.2,healthbench_consensus,94.7,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.2; HealthBench Consensus unadjusted=94.7. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.4,healthbench_consensus_length_adjusted,96.3,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.4; HealthBench Consensus length-adjusted=96.3. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.4,healthbench_consensus,96.4,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.4; HealthBench Consensus unadjusted=96.4. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.5,healthbench_consensus_length_adjusted,95.6,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.5; HealthBench Consensus length-adjusted=95.6. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.5,healthbench_consensus,95.7,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.5; HealthBench Consensus unadjusted=95.7. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.6-sol,healthbench_consensus_length_adjusted,95.5,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.6-sol; HealthBench Consensus length-adjusted=95.5. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.6-sol,healthbench_consensus,95.3,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.6-sol; HealthBench Consensus unadjusted=95.3. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.6-terra,healthbench_consensus_length_adjusted,95.1,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.6-terra; HealthBench Consensus length-adjusted=95.1. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.6-terra,healthbench_consensus,95.2,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.6-terra; HealthBench Consensus unadjusted=95.2. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.6-luna,healthbench_consensus_length_adjusted,95.1,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.6-luna; HealthBench Consensus length-adjusted=95.1. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.6-luna,healthbench_consensus,95.1,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI official HealthBench scoring"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (HealthBench length-adjusted scores): gpt-5.6-luna; HealthBench Consensus unadjusted=95.1. System Card Table 6. Adjusted and unadjusted metrics are kept as distinct benchmark identities.
gpt-5.5,aav_capsid_packaging_prediction,0.528,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI preparedness evaluation"",""tools"":""benchmark-specified""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (AAV Capsid Packaging Prediction): gpt-5.5; AAV Capsid Packaging Prediction=0.528.
gpt-5.6-sol,aav_capsid_packaging_prediction,0.529,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI preparedness evaluation"",""tools"":""benchmark-specified""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (AAV Capsid Packaging Prediction): gpt-5.6-sol; AAV Capsid Packaging Prediction=0.529.
gpt-5.4,hard_negative_protein_binding,3.5,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI preparedness evaluation"",""tools"":""benchmark-specified""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (Hard-negative protein binding prediction): gpt-5.4-thinking; Hard-negative protein binding prediction=3.5.
gpt-5.5,hard_negative_protein_binding,0.4,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI preparedness evaluation"",""tools"":""benchmark-specified""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (Hard-negative protein binding prediction): gpt-5.5; Hard-negative protein binding prediction=0.4.
gpt-5.6-sol,hard_negative_protein_binding,7.6,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI preparedness evaluation"",""tools"":""benchmark-specified""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (Hard-negative protein binding prediction): gpt-5.6-sol; Hard-negative protein binding prediction=7.6.
gpt-5.4,dna_tf_binding_design,12.82,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI preparedness evaluation"",""tools"":""benchmark-specified""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (DNA sequence design for transcription factor binding): gpt-5.4-thinking; DNA sequence design for transcription factor binding=12.82.
gpt-5.5,dna_tf_binding_design,13.82,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI preparedness evaluation"",""tools"":""benchmark-specified""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (DNA sequence design for transcription factor binding): gpt-5.5; DNA sequence design for transcription factor binding=13.82.
gpt-5.6-sol,dna_tf_binding_design,13.7,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI preparedness evaluation"",""tools"":""benchmark-specified""}",False,model_card,verified,OpenAI GPT-5.6 System Card table (DNA sequence design for transcription factor binding): gpt-5.6-sol; DNA sequence design for transcription factor binding=13.7.
gpt-5.1,first_person_fairness_harm_overall,1.28,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI First-Person Fairness Evaluation"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card figure (First-Person Fairness Evaluation): GPT-5.1 Thinking; First-Person Fairness Evaluation=1.28. All seven exact bar values are printed. Whiskers show 95% confidence intervals but endpoints are unlabeled. The card defines harm_overall as expected male-versus-female biased-answer difference based on evaluation performance divided by 10.
gpt-5.2,first_person_fairness_harm_overall,1.0,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI First-Person Fairness Evaluation"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card figure (First-Person Fairness Evaluation): GPT-5.2 Thinking; First-Person Fairness Evaluation=1.0. All seven exact bar values are printed. Whiskers show 95% confidence intervals but endpoints are unlabeled. The card defines harm_overall as expected male-versus-female biased-answer difference based on evaluation performance divided by 10.
gpt-5.4,first_person_fairness_harm_overall,0.88,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI First-Person Fairness Evaluation"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card figure (First-Person Fairness Evaluation): GPT-5.4 Thinking; First-Person Fairness Evaluation=0.88. All seven exact bar values are printed. Whiskers show 95% confidence intervals but endpoints are unlabeled. The card defines harm_overall as expected male-versus-female biased-answer difference based on evaluation performance divided by 10.
gpt-5.5,first_person_fairness_harm_overall,1.12,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI First-Person Fairness Evaluation"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card figure (First-Person Fairness Evaluation): GPT-5.5; First-Person Fairness Evaluation=1.12. All seven exact bar values are printed. Whiskers show 95% confidence intervals but endpoints are unlabeled. The card defines harm_overall as expected male-versus-female biased-answer difference based on evaluation performance divided by 10.
gpt-5.6-sol,first_person_fairness_harm_overall,0.98,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI First-Person Fairness Evaluation"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card figure (First-Person Fairness Evaluation): GPT-5.6 Sol; First-Person Fairness Evaluation=0.98. All seven exact bar values are printed. Whiskers show 95% confidence intervals but endpoints are unlabeled. The card defines harm_overall as expected male-versus-female biased-answer difference based on evaluation performance divided by 10.
gpt-5.6-terra,first_person_fairness_harm_overall,0.88,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI First-Person Fairness Evaluation"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card figure (First-Person Fairness Evaluation): GPT-5.6 Terra; First-Person Fairness Evaluation=0.88. All seven exact bar values are printed. Whiskers show 95% confidence intervals but endpoints are unlabeled. The card defines harm_overall as expected male-versus-female biased-answer difference based on evaluation performance divided by 10.
gpt-5.6-luna,first_person_fairness_harm_overall,0.61,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""OpenAI First-Person Fairness Evaluation"",""tools"":""none""}",False,model_card,verified,OpenAI GPT-5.6 System Card figure (First-Person Fairness Evaluation): GPT-5.6 Luna; First-Person Fairness Evaluation=0.61. All seven exact bar values are printed. Whiskers show 95% confidence intervals but endpoints are unlabeled. The card defines harm_overall as expected male-versus-female biased-answer difference based on evaluation performance divided by 10.
gpt-5.6-sol,frontiercyber_easy,11.0,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""Irregular agent harness"",""sampling"":""5/44"",""tools"":""agentic cyber environment""}",False,model_card,verified,OpenAI GPT-5.6 System Card figure (FrontierCyber success rates): GPT-5.6 Sol; FrontierCyber Easy=11.0. All eight plotted percentages are exact labels; caption supplies exact numerators/denominators. Results use Irregular's agent harness. Denominators vary slightly due to device availability constraints. The benchmark tests zero-day discovery and exploitation in current off-the-shelf software and hardware.
gpt-5.5,frontiercyber_easy,6.0,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""Irregular agent harness"",""sampling"":""3/44"",""tools"":""agentic cyber environment""}",False,model_card,verified,OpenAI GPT-5.6 System Card figure (FrontierCyber success rates): GPT-5.5; FrontierCyber Easy=6.0. All eight plotted percentages are exact labels; caption supplies exact numerators/denominators. Results use Irregular's agent harness. Denominators vary slightly due to device availability constraints. The benchmark tests zero-day discovery and exploitation in current off-the-shelf software and hardware.
gpt-5.6-sol,frontiercyber_medium,12.0,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""Irregular agent harness"",""sampling"":""10/77"",""tools"":""agentic cyber environment""}",False,model_card,verified,OpenAI GPT-5.6 System Card figure (FrontierCyber success rates): GPT-5.6 Sol; FrontierCyber Medium=12.0. All eight plotted percentages are exact labels; caption supplies exact numerators/denominators. Results use Irregular's agent harness. Denominators vary slightly due to device availability constraints. The benchmark tests zero-day discovery and exploitation in current off-the-shelf software and hardware.
gpt-5.5,frontiercyber_medium,6.0,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""Irregular agent harness"",""sampling"":""5/80"",""tools"":""agentic cyber environment""}",False,model_card,verified,OpenAI GPT-5.6 System Card figure (FrontierCyber success rates): GPT-5.5; FrontierCyber Medium=6.0. All eight plotted percentages are exact labels; caption supplies exact numerators/denominators. Results use Irregular's agent harness. Denominators vary slightly due to device availability constraints. The benchmark tests zero-day discovery and exploitation in current off-the-shelf software and hardware.
gpt-5.6-sol,frontiercyber_hard,5.0,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""Irregular agent harness"",""sampling"":""4/67"",""tools"":""agentic cyber environment""}",False,model_card,verified,OpenAI GPT-5.6 System Card figure (FrontierCyber success rates): GPT-5.6 Sol; FrontierCyber Hard=5.0. All eight plotted percentages are exact labels; caption supplies exact numerators/denominators. Results use Irregular's agent harness. Denominators vary slightly due to device availability constraints. The benchmark tests zero-day discovery and exploitation in current off-the-shelf software and hardware.
gpt-5.5,frontiercyber_hard,4.0,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""Irregular agent harness"",""sampling"":""3/69"",""tools"":""agentic cyber environment""}",False,model_card,verified,OpenAI GPT-5.6 System Card figure (FrontierCyber success rates): GPT-5.5; FrontierCyber Hard=4.0. All eight plotted percentages are exact labels; caption supplies exact numerators/denominators. Results use Irregular's agent harness. Denominators vary slightly due to device availability constraints. The benchmark tests zero-day discovery and exploitation in current off-the-shelf software and hardware.
gpt-5.6-sol,frontiercyber_elite,0.0,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""Irregular agent harness"",""sampling"":""0/9"",""tools"":""agentic cyber environment""}",False,model_card,verified,OpenAI GPT-5.6 System Card figure (FrontierCyber success rates): GPT-5.6 Sol; FrontierCyber Elite=0.0. All eight plotted percentages are exact labels; caption supplies exact numerators/denominators. Results use Irregular's agent harness. Denominators vary slightly due to device availability constraints. The benchmark tests zero-day discovery and exploitation in current off-the-shelf software and hardware.
gpt-5.5,frontiercyber_elite,0.0,https://deploymentsafety.openai.com/gpt-5-6,"{""effort"":""source does not state"",""harness"":""Irregular agent harness"",""sampling"":""0/12"",""tools"":""agentic cyber environment""}",False,model_card,verified,OpenAI GPT-5.6 System Card figure (FrontierCyber success rates): GPT-5.5; FrontierCyber Elite=0.0. All eight plotted percentages are exact labels; caption supplies exact numerators/denominators. Results use Irregular's agent harness. Denominators vary slightly due to device availability constraints. The benchmark tests zero-day discovery and exploitation in current off-the-shelf software and hardware.
claude-opus-4.7,long_form_virology_task_1_sequence_design,0.89,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""biology"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 2.2.4.A, page 17: Claude Opus 4.7; Long-form virology tasks [Task 1 / Sequence design]=0.89. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=biology."
claude-opus-4.8,long_form_virology_task_1_sequence_design,0.88,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""biology"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 2.2.4.A, page 17: Claude Opus 4.8; Long-form virology tasks [Task 1 / Sequence design]=0.88. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=biology."
claude-fable-5,long_form_virology_task_1_sequence_design,0.88,https://www.anthropic.com/claude-sonnet-5-system-card,"{""deployment"":""claude-mythos-5"",""effort"":""source does not state"",""harness"":""biology"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 2.2.4.A, page 17: Claude Mythos 5; Long-form virology tasks [Task 1 / Sequence design]=0.88. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=biology; deployment=claude-mythos-5."
claude-sonnet-5,long_form_virology_task_1_sequence_design,0.9,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""biology"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 2.2.4.A, page 17: Claude Sonnet 5; Long-form virology tasks [Task 1 / Sequence design]=0.90. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=biology."
claude-opus-4.7,long_form_virology_task_1_protocol_design,0.93,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""biology"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 2.2.4.A, page 17: Claude Opus 4.7; Long-form virology tasks [Task 1 / Protocol design]=0.93. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=biology."
claude-opus-4.8,long_form_virology_task_1_protocol_design,0.87,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""biology"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 2.2.4.A, page 17: Claude Opus 4.8; Long-form virology tasks [Task 1 / Protocol design]=0.87. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=biology."
claude-fable-5,long_form_virology_task_1_protocol_design,0.87,https://www.anthropic.com/claude-sonnet-5-system-card,"{""deployment"":""claude-mythos-5"",""effort"":""source does not state"",""harness"":""biology"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 2.2.4.A, page 17: Claude Mythos 5; Long-form virology tasks [Task 1 / Protocol design]=0.87. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=biology; deployment=claude-mythos-5."
claude-sonnet-5,long_form_virology_task_1_protocol_design,0.87,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""biology"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 2.2.4.A, page 17: Claude Sonnet 5; Long-form virology tasks [Task 1 / Protocol design]=0.87. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=biology."
claude-opus-4.7,long_form_virology_task_1_end_to_end,0.82,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""biology"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 2.2.4.A, page 17: Claude Opus 4.7; Long-form virology tasks [Task 1 / End-to-end]=0.82. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=biology."
claude-opus-4.8,long_form_virology_task_1_end_to_end,0.72,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""biology"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 2.2.4.A, page 17: Claude Opus 4.8; Long-form virology tasks [Task 1 / End-to-end]=0.72. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=biology."
claude-fable-5,long_form_virology_task_1_end_to_end,0.77,https://www.anthropic.com/claude-sonnet-5-system-card,"{""deployment"":""claude-mythos-5"",""effort"":""source does not state"",""harness"":""biology"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 2.2.4.A, page 17: Claude Mythos 5; Long-form virology tasks [Task 1 / End-to-end]=0.77. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=biology; deployment=claude-mythos-5."
claude-sonnet-5,long_form_virology_task_1_end_to_end,0.79,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""biology"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 2.2.4.A, page 17: Claude Sonnet 5; Long-form virology tasks [Task 1 / End-to-end]=0.79. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=biology."
claude-opus-4.7,long_form_virology_task_2_sequence_design,1.0,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""biology"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 2.2.4.A, page 17: Claude Opus 4.7; Long-form virology tasks [Task 2 / Sequence design]=1.00. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=biology."
claude-opus-4.8,long_form_virology_task_2_sequence_design,1.0,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""biology"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 2.2.4.A, page 17: Claude Opus 4.8; Long-form virology tasks [Task 2 / Sequence design]=1.00. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=biology."
claude-fable-5,long_form_virology_task_2_sequence_design,1.0,https://www.anthropic.com/claude-sonnet-5-system-card,"{""deployment"":""claude-mythos-5"",""effort"":""source does not state"",""harness"":""biology"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 2.2.4.A, page 17: Claude Mythos 5; Long-form virology tasks [Task 2 / Sequence design]=1.00. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=biology; deployment=claude-mythos-5."
claude-sonnet-5,long_form_virology_task_2_sequence_design,1.0,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""biology"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 2.2.4.A, page 17: Claude Sonnet 5; Long-form virology tasks [Task 2 / Sequence design]=1.00. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=biology."
claude-opus-4.7,long_form_virology_task_2_protocol_design,0.94,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""biology"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 2.2.4.A, page 17: Claude Opus 4.7; Long-form virology tasks [Task 2 / Protocol design]=0.94. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=biology."
claude-opus-4.8,long_form_virology_task_2_protocol_design,0.9,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""biology"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 2.2.4.A, page 17: Claude Opus 4.8; Long-form virology tasks [Task 2 / Protocol design]=0.90. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=biology."
claude-fable-5,long_form_virology_task_2_protocol_design,0.91,https://www.anthropic.com/claude-sonnet-5-system-card,"{""deployment"":""claude-mythos-5"",""effort"":""source does not state"",""harness"":""biology"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 2.2.4.A, page 17: Claude Mythos 5; Long-form virology tasks [Task 2 / Protocol design]=0.91. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=biology; deployment=claude-mythos-5."
claude-sonnet-5,long_form_virology_task_2_protocol_design,0.9,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""biology"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 2.2.4.A, page 17: Claude Sonnet 5; Long-form virology tasks [Task 2 / Protocol design]=0.90. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=biology."
claude-opus-4.7,long_form_virology_task_2_end_to_end,0.94,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""biology"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 2.2.4.A, page 17: Claude Opus 4.7; Long-form virology tasks [Task 2 / End-to-end]=0.94. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=biology."
claude-opus-4.8,long_form_virology_task_2_end_to_end,0.9,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""biology"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 2.2.4.A, page 17: Claude Opus 4.8; Long-form virology tasks [Task 2 / End-to-end]=0.90. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=biology."
claude-fable-5,long_form_virology_task_2_end_to_end,0.91,https://www.anthropic.com/claude-sonnet-5-system-card,"{""deployment"":""claude-mythos-5"",""effort"":""source does not state"",""harness"":""biology"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 2.2.4.A, page 17: Claude Mythos 5; Long-form virology tasks [Task 2 / End-to-end]=0.91. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=biology; deployment=claude-mythos-5."
claude-sonnet-5,long_form_virology_task_2_end_to_end,0.9,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""biology"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 2.2.4.A, page 17: Claude Sonnet 5; Long-form virology tasks [Task 2 / End-to-end]=0.90. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=biology."
claude-opus-4.7,ai_rd_kernel_best_speedup,371.75,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""aird"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 2.3.3.A, page 26: Claude Opus 4.7; AI R&D rule-out evaluations [Kernel task / best speedup]=371.75×. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=aird."
claude-fable-5,ai_rd_kernel_best_speedup,430.93,https://www.anthropic.com/claude-sonnet-5-system-card,"{""deployment"":""claude-mythos-5"",""effort"":""source does not state"",""harness"":""aird"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 2.3.3.A, page 26: Claude Mythos 5; AI R&D rule-out evaluations [Kernel task / best speedup]=430.93×. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=aird; deployment=claude-mythos-5."
claude-sonnet-5,ai_rd_kernel_best_speedup,284.52,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""aird"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 2.3.3.A, page 26: Claude Sonnet 5; AI R&D rule-out evaluations [Kernel task / best speedup]=284.52×. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=aird."
claude-opus-4.7,ai_rd_time_series_forecasting_mse,4.78,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""aird"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 2.3.3.A, page 26: Claude Opus 4.7; AI R&D rule-out evaluations [Time Series Forecasting / MSE, lower better]=4.78. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=aird."
claude-fable-5,ai_rd_time_series_forecasting_mse,4.51,https://www.anthropic.com/claude-sonnet-5-system-card,"{""deployment"":""claude-mythos-5"",""effort"":""source does not state"",""harness"":""aird"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 2.3.3.A, page 26: Claude Mythos 5; AI R&D rule-out evaluations [Time Series Forecasting / MSE, lower better]=4.51. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=aird; deployment=claude-mythos-5."
claude-sonnet-5,ai_rd_time_series_forecasting_mse,5.8,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""aird"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 2.3.3.A, page 26: Claude Sonnet 5; AI R&D rule-out evaluations [Time Series Forecasting / MSE, lower better]=5.80. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=aird."
claude-opus-4.7,ai_rd_llm_training_speedup,50.67,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""aird"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 2.3.3.A, page 26: Claude Opus 4.7; AI R&D rule-out evaluations [LLM training / average speedup]=50.67×. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=aird."
claude-fable-5,ai_rd_llm_training_speedup,69.61,https://www.anthropic.com/claude-sonnet-5-system-card,"{""deployment"":""claude-mythos-5"",""effort"":""source does not state"",""harness"":""aird"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 2.3.3.A, page 26: Claude Mythos 5; AI R&D rule-out evaluations [LLM training / average speedup]=69.61×. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=aird; deployment=claude-mythos-5."
claude-sonnet-5,ai_rd_llm_training_speedup,26.49,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""aird"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 2.3.3.A, page 26: Claude Sonnet 5; AI R&D rule-out evaluations [LLM training / average speedup]=26.49×. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=aird."
claude-opus-4.7,ai_rd_quadruped_rl,24.73,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""aird"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 2.3.3.A, page 26: Claude Opus 4.7; AI R&D rule-out evaluations [Quadruped RL / highest score, no hparams]=24.73. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=aird."
claude-fable-5,ai_rd_quadruped_rl,29.55,https://www.anthropic.com/claude-sonnet-5-system-card,"{""deployment"":""claude-mythos-5"",""effort"":""source does not state"",""harness"":""aird"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 2.3.3.A, page 26: Claude Mythos 5; AI R&D rule-out evaluations [Quadruped RL / highest score, no hparams]=29.55. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=aird; deployment=claude-mythos-5."
claude-sonnet-5,ai_rd_quadruped_rl,19.94,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""aird"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 2.3.3.A, page 26: Claude Sonnet 5; AI R&D rule-out evaluations [Quadruped RL / highest score, no hparams]=19.94. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=aird."
claude-opus-4.7,ai_rd_novel_compiler,70.4,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""aird"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 2.3.3.A, page 26: Claude Opus 4.7; AI R&D rule-out evaluations [Novel Compiler / pass rate complex tests]=70.4%. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=aird."
claude-fable-5,ai_rd_novel_compiler,85.3,https://www.anthropic.com/claude-sonnet-5-system-card,"{""deployment"":""claude-mythos-5"",""effort"":""source does not state"",""harness"":""aird"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 2.3.3.A, page 26: Claude Mythos 5; AI R&D rule-out evaluations [Novel Compiler / pass rate complex tests]=85.3%. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=aird; deployment=claude-mythos-5."
claude-sonnet-5,ai_rd_novel_compiler,76.1,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""aird"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 2.3.3.A, page 26: Claude Sonnet 5; AI R&D rule-out evaluations [Novel Compiler / pass rate complex tests]=76.1%. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=aird."
claude-sonnet-5,exploitbench,31.0,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""exploit"",""sampling"":""pass@1"",""tools"":""cyber execution environment""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 3.2.1.A, page 31: Claude Sonnet 5; ExploitBench [Cap%]=31. Source setting: effort=source does not state; tools=cyber execution environment; sampling=pass@1; harness=exploit."
claude-sonnet-4.6,exploitbench,24.0,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""exploit"",""sampling"":""pass@1"",""tools"":""cyber execution environment""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 3.2.1.A, page 31: Sonnet 4.6; ExploitBench [Cap%]=24. Source setting: effort=source does not state; tools=cyber execution environment; sampling=pass@1; harness=exploit."
claude-sonnet-4.6,oss_fuzz_control_flow_hijack_count,0.0,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""oss"",""sampling"":""pass@1"",""tools"":""cyber execution environment""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 3.2.2.A, page 32: Claude Sonnet 4.6; OSS-Fuzz exploit-primitive discovery [grade 1.0 count]=0. Source setting: effort=source does not state; tools=cyber execution environment; sampling=pass@1; harness=oss."
claude-sonnet-4.6,oss_fuzz_any_progress,31.6,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""oss"",""sampling"":""pass@1"",""tools"":""cyber execution environment""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 3.2.2.A, page 32: Claude Sonnet 4.6; OSS-Fuzz exploit-primitive discovery [any progress]=31.6%. Source setting: effort=source does not state; tools=cyber execution environment; sampling=pass@1; harness=oss."
claude-opus-4.8,oss_fuzz_control_flow_hijack_count,0.0,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""oss"",""sampling"":""pass@1"",""tools"":""cyber execution environment""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 3.2.2.A, page 32: Claude Opus 4.8; OSS-Fuzz exploit-primitive discovery [grade 1.0 count]=0. Source setting: effort=source does not state; tools=cyber execution environment; sampling=pass@1; harness=oss."
claude-opus-4.8,oss_fuzz_any_progress,61.5,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""oss"",""sampling"":""pass@1"",""tools"":""cyber execution environment""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 3.2.2.A, page 32: Claude Opus 4.8; OSS-Fuzz exploit-primitive discovery [any progress]=61.5%. Source setting: effort=source does not state; tools=cyber execution environment; sampling=pass@1; harness=oss."
claude-fable-5,oss_fuzz_control_flow_hijack_count,13.0,https://www.anthropic.com/claude-sonnet-5-system-card,"{""deployment"":""claude-mythos-5"",""effort"":""source does not state"",""harness"":""oss"",""sampling"":""pass@1"",""tools"":""cyber execution environment""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 3.2.2.A, page 32: Claude Mythos 5; OSS-Fuzz exploit-primitive discovery [grade 1.0 count]=13. Source setting: effort=source does not state; tools=cyber execution environment; sampling=pass@1; harness=oss; deployment=claude-mythos-5."
claude-fable-5,oss_fuzz_any_progress,80.0,https://www.anthropic.com/claude-sonnet-5-system-card,"{""deployment"":""claude-mythos-5"",""effort"":""source does not state"",""harness"":""oss"",""sampling"":""pass@1"",""tools"":""cyber execution environment""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 3.2.2.A, page 32: Claude Mythos 5; OSS-Fuzz exploit-primitive discovery [any progress]=80.0%. Source setting: effort=source does not state; tools=cyber execution environment; sampling=pass@1; harness=oss; deployment=claude-mythos-5."
claude-sonnet-5,oss_fuzz_control_flow_hijack_count,0.0,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""oss"",""sampling"":""pass@1"",""tools"":""cyber execution environment""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 3.2.2.A, page 32: Claude Sonnet 5; OSS-Fuzz exploit-primitive discovery [grade 1.0 count]=0. Source setting: effort=source does not state; tools=cyber execution environment; sampling=pass@1; harness=oss."
claude-sonnet-5,oss_fuzz_any_progress,54.5,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""oss"",""sampling"":""pass@1"",""tools"":""cyber execution environment""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 3.2.2.A, page 32: Claude Sonnet 5; OSS-Fuzz exploit-primitive discovery [any progress]=54.5%. Source setting: effort=source does not state; tools=cyber execution environment; sampling=pass@1; harness=oss."
claude-fable-5,cybergym,83.8,https://www.anthropic.com/claude-sonnet-5-system-card,"{""deployment"":""claude-mythos-5"",""effort"":""source does not state"",""harness"":""cybergym"",""sampling"":""pass@1"",""tools"":""cyber execution environment""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 3.2.3.A, page 33: Claude Mythos 5; CyberGym vulnerability discovery [targeted]=83.8%. Source setting: effort=source does not state; tools=cyber execution environment; sampling=pass@1; harness=cybergym; deployment=claude-mythos-5."
claude-sonnet-5,cybergym,52.7,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""cybergym"",""sampling"":""pass@1"",""tools"":""cyber execution environment""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 3.2.3.A, page 33: Claude Sonnet 5; CyberGym vulnerability discovery [targeted]=52.7%. Source setting: effort=source does not state; tools=cyber execution environment; sampling=pass@1; harness=cybergym."
claude-sonnet-4.6,firefox_147_exploit_development_working_exploit,0.0,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""firefox"",""sampling"":""pass@1"",""tools"":""cyber execution environment""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 3.2.4.A, page 35: Claude Sonnet 4.6; Firefox 147 exploit development [working]=0.0%. Source setting: effort=source does not state; tools=cyber execution environment; sampling=pass@1; harness=firefox."
claude-sonnet-4.6,firefox_147_exploit_development_any_success,8.8,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""firefox"",""sampling"":""pass@1"",""tools"":""cyber execution environment""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 3.2.4.A, page 35: Claude Sonnet 4.6; Firefox 147 exploit development [any success]=8.8%. Source setting: effort=source does not state; tools=cyber execution environment; sampling=pass@1; harness=firefox."
claude-opus-4.8,firefox_147_exploit_development_working_exploit,8.8,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""firefox"",""sampling"":""pass@1"",""tools"":""cyber execution environment""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 3.2.4.A, page 35: Claude Opus 4.8; Firefox 147 exploit development [working]=8.8%. Source setting: effort=source does not state; tools=cyber execution environment; sampling=pass@1; harness=firefox."
claude-opus-4.8,firefox_147_exploit_development_any_success,68.8,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""firefox"",""sampling"":""pass@1"",""tools"":""cyber execution environment""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 3.2.4.A, page 35: Claude Opus 4.8; Firefox 147 exploit development [any success]=68.8%. Source setting: effort=source does not state; tools=cyber execution environment; sampling=pass@1; harness=firefox."
claude-fable-5,firefox_147_exploit_development_working_exploit,88.4,https://www.anthropic.com/claude-sonnet-5-system-card,"{""deployment"":""claude-mythos-5"",""effort"":""source does not state"",""harness"":""firefox"",""sampling"":""pass@1"",""tools"":""cyber execution environment""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 3.2.4.A, page 35: Claude Mythos 5; Firefox 147 exploit development [working]=88.4%. Source setting: effort=source does not state; tools=cyber execution environment; sampling=pass@1; harness=firefox; deployment=claude-mythos-5."
claude-fable-5,firefox_147_exploit_development_any_success,90.0,https://www.anthropic.com/claude-sonnet-5-system-card,"{""deployment"":""claude-mythos-5"",""effort"":""source does not state"",""harness"":""firefox"",""sampling"":""pass@1"",""tools"":""cyber execution environment""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 3.2.4.A, page 35: Claude Mythos 5; Firefox 147 exploit development [any success]=90.0%. Source setting: effort=source does not state; tools=cyber execution environment; sampling=pass@1; harness=firefox; deployment=claude-mythos-5."
claude-sonnet-5,firefox_147_exploit_development_working_exploit,0.0,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""firefox"",""sampling"":""pass@1"",""tools"":""cyber execution environment""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 3.2.4.A, page 35: Claude Sonnet 5; Firefox 147 exploit development [working]=0.0%. Source setting: effort=source does not state; tools=cyber execution environment; sampling=pass@1; harness=firefox."
claude-sonnet-5,firefox_147_exploit_development_any_success,13.2,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""firefox"",""sampling"":""pass@1"",""tools"":""cyber execution environment""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 3.2.4.A, page 35: Claude Sonnet 5; Firefox 147 exploit development [any success]=13.2%. Source setting: effort=source does not state; tools=cyber execution environment; sampling=pass@1; harness=firefox."
claude-sonnet-5,swe_bench_pro,63.2,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""table8"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 8.1.A, page 115: Claude Sonnet 5; SWE-bench Pro [summary]=63.2. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=table8."
gemini-3.5-flash,swe_bench_pro,55.1,https://deepmind.google/models/model-cards/gemini-3-5-flash/,"{""effort"":""medium (runtime default)"",""harness"":""internal Antigravity for Gemini; provider/public for others"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""Gemini avg 5 runs, single attempt per run"",""snapshot"":""May 2026"",""temperature"":""default"",""tools"":""code execution""}",False,official_model_card,verified,Replace third-party provenance with Google provider-own May evaluation for SWE-Bench Pro (Public); numeric score is unchanged.
claude-sonnet-5,terminal_bench_2_1,80.4,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""table8"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 8.1.A, page 115: Claude Sonnet 5; Terminal-Bench 2.1 [summary]=80.4. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=table8."
claude-sonnet-4.6,terminal_bench_2_1,67.0,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""table8"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 8.1.A, page 115: Claude Sonnet 4.6; Terminal-Bench 2.1 [summary]=67.0. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=table8."
gemini-3.5-flash,terminal_bench_2_1,76.2,https://deepmind.google/models/model-cards/gemini-3-5-flash/,"{""effort"":""medium (runtime default)"",""harness"":""Terminus-2; Gemini self-computed, others public leaderboard"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""May 2026"",""temperature"":""default"",""tools"":""terminal agent""}",False,official_model_card,verified,Replace third-party provenance with Google provider-own May evaluation for Terminal-Bench 2.1; numeric score is unchanged.
claude-sonnet-5,browsecomp,79.4,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""high"",""tools"":""benchmark-specific""}",True,model_card,verified,Figure 8.10.2.B / Claude Sonnet 5 / high: Fixed 10M token budget with web/programmatic/code tools and context compaction.
claude-sonnet-5,hle,39.3,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""high"",""tools"":""benchmark-specific""}",True,model_card,verified,"Figure 8.10.1.B / Claude Sonnet 5 / high: 2,500 questions; 1M token cap; no tools and no context compaction."
gemini-3.5-flash,hle,40.2,https://deepmind.google/models/model-cards/gemini-3-5-flash/,"{""effort"":""medium (runtime default)"",""harness"":""Google self-computed for Gemini; provider for others"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""May 2026"",""temperature"":""default"",""tools"":""none""}",False,official_model_card,verified,Replace third-party provenance with Google provider-own May evaluation for Humanity's Last Exam: full text+MM; numeric score is unchanged.
claude-sonnet-4.6,hle_tools,46.8,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""table8"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""web search, fetch, and code execution""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 8.1.A, page 115: Claude Sonnet 4.6; Humanity’s Last Exam / with tools [summary]=46.8. Source setting: effort=max; tools=web search, fetch, and code execution; sampling=avg 5 trials unless source states otherwise; harness=table8."
claude-sonnet-5,osworld_verified,81.2,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""table8"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 8.1.A, page 115: Claude Sonnet 5; OSWorld-Verified [summary]=81.2. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=table8."
claude-sonnet-4.6,osworld_verified,78.5,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""table8"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 8.1.A, page 115: Claude Sonnet 4.6; OSWorld-Verified [summary]=78.5. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=table8."
gemini-3.5-flash,osworld_verified,78.4,https://deepmind.google/models/model-cards/gemini-3-5-flash/,"{""effort"":""medium (runtime default)"",""harness"":""OSWorld Docker, 1080p, max 100 steps"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""Gemini avg 5 runs, single attempt per run"",""snapshot"":""May 2026"",""temperature"":""default"",""tools"":""pyautogui + UI-specific function declarations""}",False,official_model_card,verified,Replace third-party provenance with Google provider-own May evaluation for OSWorld-Verified; numeric score is unchanged.
claude-sonnet-4.6,frontiercode_main_v1,15.1,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""table8"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 8.1.A, page 115: Claude Sonnet 4.6; FrontierCode v1 [summary]=15.1. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=table8."
claude-sonnet-5,automation_bench_private_heldout,13.5,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""table8"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 8.1.A, page 115: Claude Sonnet 5; AutomationBench [summary]=13.5. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=table8."
claude-sonnet-4.6,automation_bench_private_heldout,5.3,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""table8"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 8.1.A, page 115: Claude Sonnet 4.6; AutomationBench [summary]=5.3. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=table8."
claude-sonnet-5,legal_agent_benchmark_public,8.9,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""table8"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 8.1.A, page 115: Claude Sonnet 5; Legal Agent Benchmark / Full Public Set [summary]=8.9. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=table8."
claude-sonnet-4.6,legal_agent_benchmark_public,8.0,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""table8"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 8.1.A, page 115: Claude Sonnet 4.6; Legal Agent Benchmark / Full Public Set [summary]=8.0. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=table8."
claude-sonnet-5,legal_agent_benchmark_harvey_held_out,5.8,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""table8"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 8.1.A, page 115: Claude Sonnet 5; Legal Agent Benchmark / Harvey Held-Out Set [summary]=5.8. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=table8."
claude-sonnet-4.6,legal_agent_benchmark_harvey_held_out,5.4,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""table8"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 8.1.A, page 115: Claude Sonnet 4.6; Legal Agent Benchmark / Harvey Held-Out Set [summary]=5.4. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=table8."
gpt-5.5,legal_agent_benchmark_harvey_held_out,2.1,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""table8"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 8.1.A, page 115: GPT-5.5; Legal Agent Benchmark / Harvey Held-Out Set [summary]=2.1. Source setting: effort=source does not state; tools=none; sampling=avg 5 trials unless source states otherwise; harness=table8."
gemini-3.5-flash,legal_agent_benchmark_harvey_held_out,0.8,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""table8"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 8.1.A, page 115: Gemini 3.5 Flash; Legal Agent Benchmark / Harvey Held-Out Set [summary]=0.8. Source setting: effort=source does not state; tools=none; sampling=avg 5 trials unless source states otherwise; harness=table8."
claude-sonnet-5,healthbench_professional_length_adjusted,57.8,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""table8"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 8.1.A, page 115: Claude Sonnet 5; HealthBench Professional [summary]=57.8. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=table8."
claude-sonnet-4.6,healthbench_professional_length_adjusted,44.2,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""table8"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 8.1.A, page 115: Claude Sonnet 4.6; HealthBench Professional [summary]=44.2. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=table8."
gpt-5.5,frontiercode_main_v1,25.5,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""xhigh"",""harness"":""frontier"",""sampling"":""pass@1"",""tools"":""agentic benchmark harness""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.4.A, page 118: GPT-5.5; FrontierCode v1 [xhigh]=25.5%. Source setting: effort=xhigh; tools=agentic benchmark harness; sampling=pass@1; harness=frontier."
claude-opus-4.8,frontiercode_main_v1,30.3,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""high"",""harness"":""frontier"",""sampling"":""pass@1"",""tools"":""agentic benchmark harness""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.4.A, page 118: Claude Opus 4.8; FrontierCode v1 [high]=30.3%. Source setting: effort=high; tools=agentic benchmark harness; sampling=pass@1; harness=frontier."
claude-sonnet-5,frontiercode_main_v1,28.9,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""high"",""harness"":""frontier"",""sampling"":""pass@1"",""tools"":""agentic benchmark harness""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.4.A, page 118: Claude Sonnet 5; FrontierCode v1 [high]=28.9%. Source setting: effort=high; tools=agentic benchmark harness; sampling=pass@1; harness=frontier."
claude-fable-5,frontiercode_main_v1,44.7,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""frontier"",""sampling"":""pass@1"",""tools"":""agentic benchmark harness""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.4.A, page 118: Claude Fable 5; FrontierCode v1 [max]=44.7%. Source setting: effort=max; tools=agentic benchmark harness; sampling=pass@1; harness=frontier."
gpt-5.5,cursorbench_3_1,64.3,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""xhigh"",""harness"":""cursor"",""sampling"":""pass@1"",""tools"":""agentic benchmark harness""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.5.A, page 119: GPT-5.5; CursorBench [xhigh]=64.3%. Source setting: effort=xhigh; tools=agentic benchmark harness; sampling=pass@1; harness=cursor."
claude-sonnet-4.6,cursorbench_3_1,47.5,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""cursor"",""sampling"":""pass@1"",""tools"":""agentic benchmark harness""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.5.A, page 119: Claude Sonnet 4.6; CursorBench [max]=47.5%. Source setting: effort=max; tools=agentic benchmark harness; sampling=pass@1; harness=cursor."
claude-opus-4.7,cursorbench_3_1,61.6,https://cursor.com/blog/composer-2-5,"{""effort"":""xhigh"",""harness"":""Cursor"",""sampling"":""pass@1"",""source_default"":true,""tools"":""agentic""}",True,official_blog,verified,Composer 2.5 release benchmark image / CursorBench 3.1 / Opus 4.7 / xhigh: Cursor release benchmark image. Opus 4.7 and GPT-5.5 public evaluation values are self-reported.
claude-opus-4.8,cursorbench_3_1,58.4,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""high"",""harness"":""cursor"",""sampling"":""pass@1"",""tools"":""agentic benchmark harness""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.5.A, page 119: Claude Opus 4.8; CursorBench [high]=58.4%. Source setting: effort=high; tools=agentic benchmark harness; sampling=pass@1; harness=cursor."
claude-sonnet-5,cursorbench_3_1,57.0,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""high"",""harness"":""cursor"",""sampling"":""pass@1"",""tools"":""agentic benchmark harness""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.5.A, page 119: Claude Sonnet 5; CursorBench [high]=57.0%. Source setting: effort=high; tools=agentic benchmark harness; sampling=pass@1; harness=cursor."
claude-fable-5,cursorbench_3_1,72.9,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""cursor"",""sampling"":""pass@1"",""tools"":""agentic benchmark harness""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.5.A, page 119: Claude Fable 5; CursorBench [max]=72.9%. Source setting: effort=max; tools=agentic benchmark harness; sampling=pass@1; harness=cursor."
gemini-3.5-flash,arxivmath_2026_04_05,47.7,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""arxiv"",""sampling"":""4 attempts/problem"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.7.A, page 121: Gemini 3.5 Flash; ArxivMath / April+May 2026 [summary]=47.7. Source setting: effort=source does not state; tools=none; sampling=4 attempts/problem; harness=arxiv."
gpt-5.5,arxivmath_2026_04_05,72.2,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""arxiv"",""sampling"":""4 attempts/problem"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.7.A, page 121: GPT-5.5; ArxivMath / April+May 2026 [summary]=72.2. Source setting: effort=source does not state; tools=none; sampling=4 attempts/problem; harness=arxiv."
claude-opus-4.8,arxivmath_2026_04_05,71.0,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""arxiv"",""sampling"":""4 attempts/problem"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.7.A, page 121: Claude Opus 4.8; ArxivMath / April+May 2026 [summary]=71.0. Source setting: effort=max; tools=none; sampling=4 attempts/problem; harness=arxiv."
claude-fable-5,arxivmath_2026_04_05,78.6,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""arxiv"",""sampling"":""4 attempts/problem"",""tools"":""none""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.7.A, page 121: Claude Fable 5; ArxivMath / April+May 2026 [summary]=78.6. Source setting: effort=max; tools=none; sampling=4 attempts/problem; harness=arxiv."
claude-sonnet-5,arxivmath_2026_04_05,65.7,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""arxiv"",""sampling"":""4 attempts/problem"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.7.A, page 121: Claude Sonnet 5; ArxivMath / April+May 2026 [without tools]=65.7. Source setting: effort=max; tools=none; sampling=4 attempts/problem; harness=arxiv."
claude-opus-4.6,hle_tools,46.8,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""hle"",""sampling"":""pass@1"",""tools"":""python / benchmark tools""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.9.1.A, page 122: Claude Opus 4.6; Humanity’s Last Exam [with tools]=46.8%. Source setting: effort=max; tools=python / benchmark tools; sampling=pass@1; harness=hle."
claude-sonnet-5,hle_tools,52.8,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""high"",""harness"":""hle"",""sampling"":""pass@1"",""tools"":""web search, fetch, and code execution""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.9.1.B, page 123: Claude Sonnet 5; Humanity’s Last Exam [high]=52.8%. Source setting: effort=high; tools=web search, fetch, and code execution; sampling=pass@1; harness=hle."
claude-sonnet-4.6,gdp_pdf_mean_criteria_pass_rate,66.9,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""gdp"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.10.1.A, page 125: Claude Sonnet 4.6; GDP.pdf [no tools]=66.9. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=gdp."
claude-mythos,gdp_pdf_mean_criteria_pass_rate,70.6,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""gdp"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.10.1.A, page 125: Claude Mythos Preview; GDP.pdf [no tools]=70.6. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=gdp."
claude-opus-4.8,gdp_pdf_mean_criteria_pass_rate,71.2,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""gdp"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.10.1.A, page 125: Claude Opus 4.8; GDP.pdf [no tools]=71.2. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=gdp."
claude-sonnet-5,gdp_pdf_mean_criteria_pass_rate,67.5,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""gdp"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.10.1.A, page 125: Claude Sonnet 5; GDP.pdf [no tools]=67.5. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=gdp."
claude-sonnet-4.6,chartmuseum,59.3,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""vision"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.10.4.A, page 129: Claude Sonnet 4.6; ChartMuseum [no tools]=59.3. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=vision."
claude-mythos,chartmuseum,80.7,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""vision"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.10.4.A, page 129: Claude Mythos Preview; ChartMuseum [no tools]=80.7. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=vision."
claude-opus-4.8,chartmuseum,75.8,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""vision"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.10.4.A, page 129: Claude Opus 4.8; ChartMuseum [no tools]=75.8. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=vision."
claude-sonnet-5,chartmuseum,70.1,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""vision"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.10.4.A, page 129: Claude Sonnet 5; ChartMuseum [no tools]=70.1. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=vision."
claude-sonnet-4.6,charxiv_reasoning,71.6,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""vision"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.10.5.A, page 130: Claude Sonnet 4.6; CharXiv Reasoning [no tools]=71.6. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=vision."
claude-sonnet-5,charxiv_reasoning,77.0,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""vision"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.10.5.A, page 130: Claude Sonnet 5; CharXiv Reasoning [no tools]=77.0. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=vision."
claude-sonnet-4.6,officeqa,68.7,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""office"",""sampling"":""pass@1"",""tools"":""agentic benchmark harness""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.11.1.A, page 131: Claude Sonnet 4.6; OfficeQA / OfficeQA Pro [OfficeQA full]=68.7. Source setting: effort=source does not state; tools=agentic benchmark harness; sampling=pass@1; harness=office."
claude-opus-4.7,officeqa,76.3,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""office"",""sampling"":""pass@1"",""tools"":""agentic benchmark harness""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.11.1.A, page 131: Claude Opus 4.7; OfficeQA / OfficeQA Pro [OfficeQA full]=76.3. Source setting: effort=source does not state; tools=agentic benchmark harness; sampling=pass@1; harness=office."
claude-opus-4.8,officeqa,77.6,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""office"",""sampling"":""pass@1"",""tools"":""agentic benchmark harness""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.11.1.A, page 131: Claude Opus 4.8; OfficeQA / OfficeQA Pro [OfficeQA full]=77.6. Source setting: effort=source does not state; tools=agentic benchmark harness; sampling=pass@1; harness=office."
claude-sonnet-5,officeqa,73.3,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""office"",""sampling"":""pass@1"",""tools"":""agentic benchmark harness""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.11.1.A, page 131: Claude Sonnet 5; OfficeQA / OfficeQA Pro [OfficeQA full]=73.3. Source setting: effort=source does not state; tools=agentic benchmark harness; sampling=pass@1; harness=office."
claude-sonnet-4.6,officeqa_pro,53.4,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""office"",""sampling"":""pass@1"",""tools"":""agentic benchmark harness""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.11.1.A, page 131: Claude Sonnet 4.6; OfficeQA / OfficeQA Pro [OfficeQA Pro]=53.4. Source setting: effort=source does not state; tools=agentic benchmark harness; sampling=pass@1; harness=office."
claude-sonnet-5,officeqa_pro,59.4,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""office"",""sampling"":""pass@1"",""tools"":""agentic benchmark harness""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.11.1.A, page 131: Claude Sonnet 5; OfficeQA / OfficeQA Pro [OfficeQA Pro]=59.4. Source setting: effort=source does not state; tools=agentic benchmark harness; sampling=pass@1; harness=office."
claude-sonnet-4.6,real_world_finance_v2_elo,1000.0,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""finance"",""sampling"":""pass@1"",""tools"":""agentic benchmark harness""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.11.2.A, page 133: Sonnet 4.6; Real-World Finance v2 [summary]=1000. Source setting: effort=max; tools=agentic benchmark harness; sampling=pass@1; harness=finance."
claude-opus-4.7,real_world_finance_v2_elo,1207.0,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""finance"",""sampling"":""pass@1"",""tools"":""agentic benchmark harness""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.11.2.A, page 133: Opus 4.7; Real-World Finance v2 [summary]=1207. Source setting: effort=max; tools=agentic benchmark harness; sampling=pass@1; harness=finance."
claude-opus-4.8,real_world_finance_v2_elo,1222.0,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""finance"",""sampling"":""pass@1"",""tools"":""agentic benchmark harness""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.11.2.A, page 133: Opus 4.8; Real-World Finance v2 [summary]=1222. Source setting: effort=max; tools=agentic benchmark harness; sampling=pass@1; harness=finance."
claude-fable-5,real_world_finance_v2_elo,1374.0,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""finance"",""sampling"":""pass@1"",""tools"":""agentic benchmark harness""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.11.2.A, page 133: Fable 5; Real-World Finance v2 [summary]=1374. Source setting: effort=max; tools=agentic benchmark harness; sampling=pass@1; harness=finance."
claude-sonnet-5,real_world_finance_v2_elo,1219.0,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""finance"",""sampling"":""pass@1"",""tools"":""agentic benchmark harness""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.11.2.A, page 133: Sonnet 5; Real-World Finance v2 [summary]=1219. Source setting: effort=max; tools=agentic benchmark harness; sampling=pass@1; harness=finance."
claude-sonnet-5,gdpval_aa_elo,1618.0,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""gdpval"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Section 8.11.4, page 134: Claude Sonnet 5; GDPval-AA v2 [summary]=1618. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=gdpval."
claude-sonnet-5,aa_briefcase_elo,1393.0,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""briefcase"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Section 8.11.7, page 137: Claude Sonnet 5; AA-Briefcase [summary]=1393. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=briefcase."
claude-sonnet-5,toolathlon,54.3,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""toolathlon"",""sampling"":""3 trials/task"",""tools"":""agentic benchmark harness""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 8.11.5.A, page 135: Claude Sonnet 5; Toolathlon [Pass@1]=54.3. Source setting: effort=max; tools=agentic benchmark harness; sampling=3 trials/task; harness=toolathlon."
claude-opus-4.7,toolathlon,59.3,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""toolathlon"",""sampling"":""3 trials/task"",""tools"":""agentic benchmark harness""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 8.11.5.A, page 135: Claude Opus 4.7; Toolathlon [Pass@1]=59.3. Source setting: effort=max; tools=agentic benchmark harness; sampling=3 trials/task; harness=toolathlon."
claude-sonnet-4.6,toolathlon,49.4,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""toolathlon"",""sampling"":""3 trials/task"",""tools"":""agentic benchmark harness""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Table 8.11.5.A, page 135: Claude Sonnet 4.6; Toolathlon [Pass@1]=49.4. Source setting: effort=max; tools=agentic benchmark harness; sampling=3 trials/task; harness=toolathlon."
gemini-3.5-flash,automation_bench_private_heldout,12.6,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""high"",""harness"":""automation"",""sampling"":""pass@1"",""tools"":""agentic benchmark harness""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.11.6.A, page 136: Gemini 3.5 Flash; AutomationBench [high]=12.6%. Source setting: effort=high; tools=agentic benchmark harness; sampling=pass@1; harness=automation."
gpt-5.5,automation_bench_private_heldout,12.9,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""xhigh"",""harness"":""automation"",""sampling"":""pass@1"",""tools"":""agentic benchmark harness""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.11.6.A, page 136: GPT-5.5; AutomationBench [xhigh]=12.9%. Source setting: effort=xhigh; tools=agentic benchmark harness; sampling=pass@1; harness=automation."
claude-opus-4.8,automation_bench_private_heldout,15.5,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""automation"",""sampling"":""pass@1"",""tools"":""agentic benchmark harness""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.11.6.A, page 136: Claude Opus 4.8; AutomationBench [max]=15.5%. Source setting: effort=max; tools=agentic benchmark harness; sampling=pass@1; harness=automation."
claude-fable-5,automation_bench_private_heldout,17.4,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""automation"",""sampling"":""pass@1"",""tools"":""agentic benchmark harness""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.11.6.A, page 136: Claude Fable 5; AutomationBench [max]=17.4%. Source setting: effort=max; tools=agentic benchmark harness; sampling=pass@1; harness=automation."
claude-sonnet-4.6,healthbench_length_adjusted,49.9,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""health"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.12.1.A, page 138: Claude Sonnet 4.6; HealthBench [summary]=49.9%. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=health."
claude-mythos,healthbench_length_adjusted,61.1,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""health"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.12.1.A, page 138: Claude Mythos Preview; HealthBench [summary]=61.1%. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=health."
claude-opus-4.8,healthbench_length_adjusted,59.3,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""health"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.12.1.A, page 138: Claude Opus 4.8; HealthBench [summary]=59.3%. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=health."
claude-sonnet-5,healthbench_length_adjusted,58.7,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""health"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.12.1.A, page 138: Claude Sonnet 5; HealthBench [summary]=58.7%. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=health."
claude-mythos,healthbench_professional_length_adjusted,64.0,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""health"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.12.2.A, page 139: Claude Mythos Preview; HealthBench Professional [summary]=64.0%. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=health."
claude-sonnet-4.6,global_mmlu,88.4,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""gmmlu"",""sampling"":""pass@1"",""tools"":""none""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.13.1.A, page 140: Claude Sonnet 4.6; GMMLU [summary]=88.4. Source setting: effort=max; tools=none; sampling=pass@1; harness=gmmlu."
claude-mythos,global_mmlu,92.9,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""gmmlu"",""sampling"":""pass@1"",""tools"":""none""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.13.1.A, page 140: Claude Mythos Preview; GMMLU [summary]=92.9. Source setting: effort=max; tools=none; sampling=pass@1; harness=gmmlu."
claude-opus-4.8,global_mmlu,90.9,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""gmmlu"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.13.1.A, page 140: Claude Opus 4.8; GMMLU [summary]=90.9. Source setting: effort=max; tools=none; sampling=pass@1; harness=gmmlu."
claude-sonnet-5,global_mmlu,89.0,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""gmmlu"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.13.1.A, page 140: Claude Sonnet 5; GMMLU [summary]=89.0. Source setting: effort=max; tools=none; sampling=pass@1; harness=gmmlu."
claude-sonnet-4.6,milu,89.4,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""multilingual"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.13.2.A, page 141: Claude Sonnet 4.6; MILU [summary]=89.4. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=multilingual."
claude-mythos,milu,92.7,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""multilingual"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.13.2.A, page 141: Claude Mythos Preview; MILU [summary]=92.7. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=multilingual."
claude-opus-4.8,milu,91.1,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""multilingual"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.13.2.A, page 141: Claude Opus 4.8; MILU [summary]=91.1. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=multilingual."
claude-sonnet-5,milu,89.3,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""multilingual"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.13.2.A, page 141: Claude Sonnet 5; MILU [summary]=89.3. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=multilingual."
claude-sonnet-4.6,include_base_44,85.7,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""multilingual"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.13.3.A, page 142: Claude Sonnet 4.6; INCLUDE [summary]=85.7. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=multilingual."
claude-mythos,include_base_44,90.1,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""multilingual"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.13.3.A, page 142: Claude Mythos Preview; INCLUDE [summary]=90.1. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=multilingual."
claude-opus-4.8,include_base_44,88.5,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""multilingual"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.13.3.A, page 142: Claude Opus 4.8; INCLUDE [summary]=88.5. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=multilingual."
claude-sonnet-5,include_base_44,86.5,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""multilingual"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.13.3.A, page 142: Claude Sonnet 5; INCLUDE [summary]=86.5. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=multilingual."
claude-sonnet-4.6,biomysterybench_human_solvable,0.78,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""life"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.14.6.A, page 144: Claude Sonnet 4.6; BioMysteryBench / Human solvable [summary]=0.78. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=life."
claude-opus-4.8,biomysterybench_human_solvable,0.8,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""life"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.14.6.A, page 144: Claude Opus 4.8; BioMysteryBench / Human solvable [summary]=0.80. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=life."
claude-fable-5,biomysterybench_human_solvable,0.84,https://www.anthropic.com/claude-sonnet-5-system-card,"{""deployment"":""claude-mythos-5"",""effort"":""source does not state"",""harness"":""life"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.14.6.A, page 144: Claude Mythos 5; BioMysteryBench / Human solvable [summary]=0.84. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=life; deployment=claude-mythos-5."
claude-sonnet-5,biomysterybench_human_solvable,0.82,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""life"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.14.6.A, page 144: Claude Sonnet 5; BioMysteryBench / Human solvable [summary]=0.82. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=life."
claude-sonnet-4.6,biomysterybench_human_difficult,0.31,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""life"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.14.6.A, page 144: Claude Sonnet 4.6; BioMysteryBench / Human difficult [summary]=0.31. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=life."
claude-opus-4.8,biomysterybench_human_difficult,0.4,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""life"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.14.6.A, page 144: Claude Opus 4.8; BioMysteryBench / Human difficult [summary]=0.40. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=life."
claude-fable-5,biomysterybench_human_difficult,0.46,https://www.anthropic.com/claude-sonnet-5-system-card,"{""deployment"":""claude-mythos-5"",""effort"":""source does not state"",""harness"":""life"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.14.6.A, page 144: Claude Mythos 5; BioMysteryBench / Human difficult [summary]=0.46. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=life; deployment=claude-mythos-5."
claude-sonnet-5,biomysterybench_human_difficult,0.35,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""life"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.14.6.A, page 144: Claude Sonnet 5; BioMysteryBench / Human difficult [summary]=0.35. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=life."
claude-sonnet-4.6,spatialbench_verified,0.6,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""life"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.14.6.A, page 144: Claude Sonnet 4.6; LatchBio Bioinformatics / SpatialBench Verified [summary]=0.60. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=life."
claude-opus-4.8,spatialbench_verified,0.67,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""life"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.14.6.A, page 144: Claude Opus 4.8; LatchBio Bioinformatics / SpatialBench Verified [summary]=0.67. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=life."
claude-fable-5,spatialbench_verified,0.7,https://www.anthropic.com/claude-sonnet-5-system-card,"{""deployment"":""claude-mythos-5"",""effort"":""source does not state"",""harness"":""life"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.14.6.A, page 144: Claude Mythos 5; LatchBio Bioinformatics / SpatialBench Verified [summary]=0.70. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=life; deployment=claude-mythos-5."
claude-sonnet-5,spatialbench_verified,0.7,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""life"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.14.6.A, page 144: Claude Sonnet 5; LatchBio Bioinformatics / SpatialBench Verified [summary]=0.70. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=life."
claude-sonnet-4.6,singlecellbench,0.5,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""life"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.14.6.A, page 144: Claude Sonnet 4.6; LatchBio Bioinformatics / SingleCellBench [summary]=0.50. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=life."
claude-opus-4.8,singlecellbench,0.58,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""life"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.14.6.A, page 144: Claude Opus 4.8; LatchBio Bioinformatics / SingleCellBench [summary]=0.58. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=life."
claude-fable-5,singlecellbench,0.59,https://www.anthropic.com/claude-sonnet-5-system-card,"{""deployment"":""claude-mythos-5"",""effort"":""source does not state"",""harness"":""life"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.14.6.A, page 144: Claude Mythos 5; LatchBio Bioinformatics / SingleCellBench [summary]=0.59. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=life; deployment=claude-mythos-5."
claude-sonnet-5,singlecellbench,0.56,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""life"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.14.6.A, page 144: Claude Sonnet 5; LatchBio Bioinformatics / SingleCellBench [summary]=0.56. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=life."
claude-sonnet-4.6,structural_biology_open_ended_internal,0.32,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""life"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.14.6.A, page 144: Claude Sonnet 4.6; Structural biology open-ended [summary]=0.32. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=life."
claude-opus-4.8,structural_biology_open_ended_internal,0.77,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""life"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.14.6.A, page 144: Claude Opus 4.8; Structural biology open-ended [summary]=0.77. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=life."
claude-fable-5,structural_biology_open_ended_internal,0.87,https://www.anthropic.com/claude-sonnet-5-system-card,"{""deployment"":""claude-mythos-5"",""effort"":""source does not state"",""harness"":""life"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.14.6.A, page 144: Claude Mythos 5; Structural biology open-ended [summary]=0.87. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=life; deployment=claude-mythos-5."
claude-sonnet-5,structural_biology_open_ended_internal,0.69,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""life"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.14.6.A, page 144: Claude Sonnet 5; Structural biology open-ended [summary]=0.69. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=life."
claude-sonnet-4.6,proteingym_hard,0.35,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""life"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.14.6.A, page 144: Claude Sonnet 4.6; ProteinGym Hard [summary]=0.35. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=life."
claude-opus-4.8,proteingym_hard,0.4,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""life"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.14.6.A, page 144: Claude Opus 4.8; ProteinGym Hard [summary]=0.40. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=life."
claude-fable-5,proteingym_hard,0.45,https://www.anthropic.com/claude-sonnet-5-system-card,"{""deployment"":""claude-mythos-5"",""effort"":""source does not state"",""harness"":""life"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.14.6.A, page 144: Claude Mythos 5; ProteinGym Hard [summary]=0.45. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=life; deployment=claude-mythos-5."
claude-sonnet-5,proteingym_hard,0.37,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""life"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.14.6.A, page 144: Claude Sonnet 5; ProteinGym Hard [summary]=0.37. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=life."
claude-sonnet-4.6,organic_chemistry_internal,0.56,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""life"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.14.6.A, page 144: Claude Sonnet 4.6; Organic chemistry [summary]=0.56. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=life."
claude-opus-4.8,organic_chemistry_internal,0.86,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""life"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.14.6.A, page 144: Claude Opus 4.8; Organic chemistry [summary]=0.86. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=life."
claude-fable-5,organic_chemistry_internal,0.9,https://www.anthropic.com/claude-sonnet-5-system-card,"{""deployment"":""claude-mythos-5"",""effort"":""source does not state"",""harness"":""life"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.14.6.A, page 144: Claude Mythos 5; Organic chemistry [summary]=0.90. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=life; deployment=claude-mythos-5."
claude-sonnet-5,organic_chemistry_internal,0.81,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""life"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.14.6.A, page 144: Claude Sonnet 5; Organic chemistry [summary]=0.81. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=life."
claude-sonnet-4.6,protocol_troubleshooting_internal,0.42,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""life"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.14.6.A, page 144: Claude Sonnet 4.6; Protocol troubleshooting [summary]=0.42. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=life."
claude-opus-4.8,protocol_troubleshooting_internal,0.6,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""life"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.14.6.A, page 144: Claude Opus 4.8; Protocol troubleshooting [summary]=0.60. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=life."
claude-fable-5,protocol_troubleshooting_internal,0.67,https://www.anthropic.com/claude-sonnet-5-system-card,"{""deployment"":""claude-mythos-5"",""effort"":""source does not state"",""harness"":""life"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.14.6.A, page 144: Claude Mythos 5; Protocol troubleshooting [summary]=0.67. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=life; deployment=claude-mythos-5."
claude-sonnet-5,protocol_troubleshooting_internal,0.62,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""source does not state"",""harness"":""life"",""sampling"":""pass@1"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Figure 8.14.6.A, page 144: Claude Sonnet 5; Protocol troubleshooting [summary]=0.62. Source setting: effort=source does not state; tools=none; sampling=pass@1; harness=life."
claude-sonnet-5,swe_bench_verified,85.2,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""standard"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Section 8.2, page 116: Claude Sonnet 5; SWE-bench Verified / Multilingual / Multimodal [Verified]=85.2%. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=standard."
claude-sonnet-5,swe_bench_multilingual,78.3,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""standard"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Section 8.2, page 116: Claude Sonnet 5; SWE-bench Verified / Multilingual / Multimodal [Multilingual]=78.3%. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=standard."
claude-sonnet-5,swe_bench_multimodal,28.1,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""max"",""harness"":""standard"",""sampling"":""avg 5 trials unless source states otherwise"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Section 8.2, page 116: Claude Sonnet 5; SWE-bench Verified / Multilingual / Multimodal [Multimodal]=28.1%. Source setting: effort=max; tools=none; sampling=avg 5 trials unless source states otherwise; harness=standard."
claude-sonnet-5,usamo_2026,79.5,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""high"",""harness"":""usamo"",""sampling"":""10 attempts/problem"",""tools"":""none""}",True,model_card,verified,"Anthropic Claude Sonnet 5 System Card Section 8.6, page 120: Claude Sonnet 5; USAMO 2026 [summary]=79.5%. Source setting: effort=high; tools=none; sampling=10 attempts/problem; harness=usamo."
claude-sonnet-4.6,usamo_2026,55.0,https://www.anthropic.com/claude-sonnet-5-system-card,"{""effort"":""high"",""harness"":""usamo"",""sampling"":""10 attempts/problem"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Section 8.6, page 120: Claude Sonnet 4.6; USAMO 2026 [summary]=55.0%. Source setting: effort=high; tools=none; sampling=10 attempts/problem; harness=usamo."
claude-fable-5,usamo_2026,99.8,https://www.anthropic.com/claude-sonnet-5-system-card,"{""deployment"":""claude-mythos-5"",""effort"":""high"",""harness"":""usamo"",""sampling"":""10 attempts/problem"",""tools"":""none""}",False,model_card,verified,"Anthropic Claude Sonnet 5 System Card Section 8.6, page 120: Claude Mythos 5; USAMO 2026 [summary]=99.8%. Source setting: effort=high; tools=none; sampling=10 attempts/problem; harness=usamo; deployment=claude-mythos-5."
claude-opus-5,swe_bench_pro,79.2,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""harness"":""official or cited leaderboard""}",False,model_card,verified,"Table 8.1.A / SWE-bench Pro: Unless noted, Anthropic values use adaptive thinking at max effort, default sampling, mean over five trials. Table 8.1.A card summary."
claude-opus-5,swe_bench_multilingual,89.5,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""harness"":""official or cited leaderboard""}",False,model_card,verified,"Table 8.1.A / SWE-bench Multilingual: Unless noted, Anthropic values use adaptive thinking at max effort, default sampling, mean over five trials. Table 8.1.A card summary."
claude-fable-5,swe_bench_multilingual,86.6,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""harness"":""official or cited leaderboard""}",True,model_card,verified,"Table 8.1.A / SWE-bench Multilingual: Unless noted, Anthropic values use adaptive thinking at max effort, default sampling, mean over five trials. Table 8.1.A card summary."
claude-opus-5,swe_bench_multimodal,59.4,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""harness"":""official or cited leaderboard""}",False,model_card,verified,"Table 8.1.A / SWE-bench Multimodal: Unless noted, Anthropic values use adaptive thinking at max effort, default sampling, mean over five trials. Table 8.1.A card summary."
claude-opus-4.8,swe_bench_multimodal,38.4,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""harness"":""official or cited leaderboard""}",False,model_card,verified,"Table 8.1.A / SWE-bench Multimodal: Unless noted, Anthropic values use adaptive thinking at max effort, default sampling, mean over five trials. Table 8.1.A card summary."
claude-fable-5,swe_bench_multimodal,54.1,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""harness"":""official or cited leaderboard""}",True,model_card,verified,"Table 8.1.A / SWE-bench Multimodal: Unless noted, Anthropic values use adaptive thinking at max effort, default sampling, mean over five trials. Table 8.1.A card summary."
claude-opus-5,healthbench_professional_length_adjusted,59.8,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""harness"":""official or cited leaderboard""}",False,model_card,verified,"Table 8.1.A / HealthBench Professional Length-Adjusted: Unless noted, Anthropic values use adaptive thinking at max effort, default sampling, mean over five trials. Table 8.1.A card summary."
gpt-5.6-sol,automation_bench_private_heldout,18.1,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""harness"":""official or cited leaderboard""}",True,model_card,verified,"Table 8.1.A / AutomationBench Private Held-Out: Unless noted, Anthropic values use adaptive thinking at max effort, default sampling, mean over five trials. Table 8.1.A card summary."
claude-opus-4.8,arc_agi_1,92.5,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""harness"":""official or cited leaderboard""}",False,model_card,verified,"Table 8.1.A / ARC-AGI-1: Unless noted, Anthropic values use adaptive thinking at max effort, default sampling, mean over five trials. Table 8.1.A card summary."
gpt-5.6-sol,arc_agi_1,97.5,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""xhigh"",""harness"":""official or cited leaderboard""}",False,model_card,verified,"Table 8.1.A / ARC-AGI-1: Unless noted, Anthropic values use adaptive thinking at max effort, default sampling, mean over five trials. Table 8.1.A card summary."
claude-opus-4.8,arc_agi_2,72.1,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""harness"":""official or cited leaderboard""}",False,model_card,verified,"Table 8.1.A / ARC-AGI-2: Unless noted, Anthropic values use adaptive thinking at max effort, default sampling, mean over five trials. Table 8.1.A card summary."
gpt-5.6-sol,arc_agi_2,92.5,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""harness"":""official or cited leaderboard""}",True,model_card,verified,"Table 8.1.A / ARC-AGI-2: Unless noted, Anthropic values use adaptive thinking at max effort, default sampling, mean over five trials. Table 8.1.A card summary."
claude-opus-5,swe_bench_verified,96.0,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""harness"":""official""}",False,model_card,verified,SWE-bench Verified narrative: 500 problems; average over five trials.
claude-sonnet-5,deep_swe_v1_1,48.2,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""high"",""tools"":""benchmark-specific""}",True,model_card,verified,Figure 8.3.A / Claude Sonnet 5 / high: 113 tasks; five trials.
claude-opus-5,deep_swe_v1_1,68.0,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""high"",""tools"":""benchmark-specific""}",True,model_card,verified,Figure 8.3.A / Claude Opus 5 / high: 113 tasks; five trials.
claude-opus-5,hle_tools,63.2,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""high"",""tools"":""benchmark-specific""}",True,model_card,verified,"Figure 8.10.1.A / Claude Opus 5 / high: 2,500 questions; 1M token cap; web/programmatic/code tools; no context compaction."
claude-opus-5,hle,56.0,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""high"",""tools"":""benchmark-specific""}",True,model_card,verified,"Figure 8.10.1.B / Claude Opus 5 / high: 2,500 questions; 1M token cap; no tools and no context compaction."
claude-opus-5,browsecomp,90.2,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""high"",""tools"":""benchmark-specific""}",True,model_card,verified,Figure 8.10.2.B / Claude Opus 5 / high: Fixed 10M token budget with web/programmatic/code tools and context compaction.
claude-sonnet-5,deepsearchqa_f1,89.4,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""high"",""tools"":""benchmark-specific""}",True,model_card,verified,Figure 8.10.3.A / Claude Sonnet 5 / high: Mean F1 over 900 prompts in 17 fields; web tools; nominal 1M token budget.
claude-opus-5,deepsearchqa_f1,94.1,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""high"",""tools"":""benchmark-specific""}",True,model_card,verified,Figure 8.10.3.A / Claude Opus 5 / high: Mean F1 over 900 prompts in 17 fields; web tools; nominal 1M token budget.
claude-opus-4.8,draco,76.3,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""high"",""tools"":""benchmark-specific""}",True,model_card,verified,Figure 8.10.4.A / Claude Opus 4.8 / high: 100 curated agentic data-analysis tasks; Opus 4.6 judge; five grading runs.
claude-fable-5,draco,87.7,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""benchmark-specific""}",True,model_card,verified,Figure 8.10.4.A / Claude Fable 5 / max: 100 curated agentic data-analysis tasks; Opus 4.6 judge; five grading runs.
claude-sonnet-5,draco,76.4,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""high"",""tools"":""benchmark-specific""}",True,model_card,verified,Figure 8.10.4.A / Claude Sonnet 5 / high: 100 curated agentic data-analysis tasks; Opus 4.6 judge; five grading runs.
claude-opus-5,draco,87.0,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""high"",""tools"":""benchmark-specific""}",True,model_card,verified,Figure 8.10.4.A / Claude Opus 5 / high: 100 curated agentic data-analysis tasks; Opus 4.6 judge; five grading runs.
gpt-5.6-sol,frontiercode_main_v1_1,47.5,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""harness"":""Cognition""}",True,model_card,verified,Figure 8.4.A / GPT 5.6 Sol / max: Five effort settings; score is averaged over five trials.
claude-opus-4.8,frontiercode_main_v1_1,41.0,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""high"",""harness"":""Cognition""}",True,model_card,verified,Figure 8.4.A / Claude Opus 4.8 / high: Five effort settings; score is averaged over five trials.
claude-fable-5,frontiercode_main_v1_1,51.6,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""harness"":""Cognition""}",True,model_card,verified,Figure 8.4.A / Claude Fable 5 / max: Five effort settings; score is averaged over five trials.
claude-opus-5,frontiercode_main_v1_1,48.0,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""high"",""harness"":""Cognition""}",True,model_card,verified,Figure 8.4.A / Claude Opus 5 / high: Five effort settings; score is averaged over five trials.
gpt-5.6-sol,frontiercode_extended_v1_1,60.6,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""harness"":""Cognition""}",True,model_card,verified,Figure 8.4.B / GPT 5.6 Sol / max: Five effort settings; score is averaged over five trials.
claude-opus-4.8,frontiercode_extended_v1_1,55.6,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""high"",""harness"":""Cognition""}",True,model_card,verified,Figure 8.4.B / Claude Opus 4.8 / high: Five effort settings; score is averaged over five trials.
claude-fable-5,frontiercode_extended_v1_1,64.9,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max, with fallback"",""harness"":""FrontierCode v1.1 Extended harness"",""judge"":""maintainer-authored rubrics, verifiers, and hard blockers"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""fixed coding-agent scaffold""}",True,model_card,verified,"card46 FrontierCode v1.1 Extended Extended score (%); source effort=max, with fallback; benchmark protocol matches canonical=True. Unknown source fields remain unknown."
claude-opus-5,frontiercode_extended_v1_1,58.5,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""high"",""harness"":""Cognition""}",True,model_card,verified,Figure 8.4.B / Claude Opus 5 / high: Five effort settings; score is averaged over five trials.
claude-opus-5,frontierbench_v0_1,39.0,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""high"",""harness"":""mini-SWE-agent on GKE""}",True,model_card,verified,FrontierBench narrative / Claude Opus 5 / high: 74 tasks; five attempts per task.
claude-fable-5,frontierbench_v0_1,33.7,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""harness"":""mini-SWE-agent on GKE""}",True,model_card,verified,FrontierBench narrative / Claude Fable 5: 74 tasks; five attempts per task.
claude-sonnet-5,frontierbench_v0_1,17.0,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""harness"":""mini-SWE-agent on GKE""}",False,model_card,verified,FrontierBench narrative / Claude Sonnet 5: 74 tasks; five attempts per task.
claude-opus-4.8,frontierbench_v0_1,18.7,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""harness"":""mini-SWE-agent on GKE""}",False,model_card,verified,FrontierBench narrative / Claude Opus 4.8: 74 tasks; five attempts per task.
gpt-5.6-sol,frontierbench_v0_1,37.5,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""harness"":""mini-SWE-agent on GKE""}",True,model_card,verified,FrontierBench narrative / GPT 5.6 Sol: 74 tasks; five attempts per task.
claude-mythos,riemannbench_no_tools,53.0,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""none""}",False,model_card,verified,Figure 8.7.A / Claude Mythos Preview / no tools: Private 25-problem benchmark; corrected references/grading; four attempts per problem.
claude-mythos,riemannbench_with_tools,67.0,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""benchmark-specified""}",False,model_card,verified,Figure 8.7.A / Claude Mythos Preview / with tools: Private 25-problem benchmark; corrected references/grading; four attempts per problem.
claude-opus-4.8,riemannbench_no_tools,48.0,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""none""}",False,model_card,verified,Figure 8.7.A / Claude Opus 4.8 / no tools: Private 25-problem benchmark; corrected references/grading; four attempts per problem.
claude-opus-4.8,riemannbench_with_tools,66.0,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""benchmark-specified""}",False,model_card,verified,Figure 8.7.A / Claude Opus 4.8 / with tools: Private 25-problem benchmark; corrected references/grading; four attempts per problem.
claude-opus-5,riemannbench_no_tools,60.0,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""none""}",False,model_card,verified,Figure 8.7.A / Claude Opus 5 / no tools: Private 25-problem benchmark; corrected references/grading; four attempts per problem.
claude-opus-5,riemannbench_with_tools,79.0,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""benchmark-specified""}",False,model_card,verified,Figure 8.7.A / Claude Opus 5 / with tools: Private 25-problem benchmark; corrected references/grading; four attempts per problem.
grok-4.5,arxivmath_2026_06_no_tools,65.99,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""source does not state"",""tools"":""none""}",False,model_card,verified,Figure 8.8.A / Grok 4.5 / no tools: 49 June 2026 problems; Anthropic internal values average four runs.
gemini-3.1-pro,arxivmath_2026_06_no_tools,65.99,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""source does not state"",""tools"":""none""}",False,model_card,verified,Figure 8.8.A / Gemini 3.1 Pro Preview / no tools: 49 June 2026 problems; Anthropic internal values average four runs.
gpt-5.6-sol,arxivmath_2026_06_no_tools,86.73,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""source does not state"",""tools"":""none""}",False,model_card,verified,Figure 8.8.A / GPT-5.6-Sol / no tools: 49 June 2026 problems; Anthropic internal values average four runs.
claude-opus-4.8,arxivmath_2026_06_no_tools,82.65,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""none""}",False,model_card,verified,Figure 8.8.A / Claude Opus 4.8 / no tools: 49 June 2026 problems; Anthropic internal values average four runs.
claude-opus-5,arxivmath_2026_06_no_tools,90.82,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""none""}",False,model_card,verified,Figure 8.8.A / Claude Opus 5 / no tools: 49 June 2026 problems; Anthropic internal values average four runs.
claude-opus-5,arxivmath_2026_06_with_tools,91.33,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""benchmark-specified""}",False,model_card,verified,Figure 8.8.A / Claude Opus 5 / with tools: 49 June 2026 problems; four runs per problem.
gemini-3.5-flash,chartography_no_tools,35.9,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""none""}",False,model_card,verified,Figure 8.12.1.A / Gemini 3.5 Flash / no tools: 100 specialized chart types; five runs; 95% CI.
gpt-5.6-sol,chartography_no_tools,45.0,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""none""}",True,model_card,verified,Figure 8.12.1.A / GPT 5.6 Sol / no tools: 100 specialized chart types; five runs; 95% CI.
claude-opus-4.8,chartography_no_tools,17.0,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""none""}",False,model_card,verified,Figure 8.12.1.A / Claude Opus 4.8 / no tools: 100 specialized chart types; five runs; 95% CI.
claude-opus-4.8,chartography_with_tools,75.0,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""benchmark-specified""}",False,model_card,verified,Figure 8.12.1.A / Claude Opus 4.8 / tools: 100 specialized chart types; five runs; 95% CI.
claude-sonnet-5,chartography_no_tools,16.0,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""none""}",False,model_card,verified,Figure 8.12.1.A / Claude Sonnet 5 / no tools: 100 specialized chart types; five runs; 95% CI.
claude-sonnet-5,chartography_with_tools,71.8,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""benchmark-specified""}",False,model_card,verified,Figure 8.12.1.A / Claude Sonnet 5 / tools: 100 specialized chart types; five runs; 95% CI.
claude-opus-5,chartography_no_tools,29.6,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""none""}",False,model_card,verified,Figure 8.12.1.A / Claude Opus 5 / no tools: 100 specialized chart types; five runs; 95% CI.
claude-opus-5,chartography_with_tools,83.0,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""benchmark-specified""}",False,model_card,verified,Figure 8.12.1.A / Claude Opus 5 / tools: 100 specialized chart types; five runs; 95% CI.
claude-opus-5,gdp_pdf_mean_criteria_pass_rate,83.4,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""harness"":""corrected full 100-prompt run"",""tools"":""none""}",False,model_card,verified,"Figure 8.12.4.A narrative / Claude Opus 5 / no tools: 100 prompts, 10 domains, corrected PDF truncation bug, Opus 4.7 judge."
claude-opus-5,officeqa,78.1,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max""}",False,model_card,verified,OfficeQA narrative / Claude Opus 5: Public Messages API for Opus 5; production safeguards and Opus 4.8 classifier fallback.
claude-opus-5,officeqa_pro,66.9,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max""}",False,model_card,verified,OfficeQA Pro narrative / Claude Opus 5: 133-question subset; public Messages API for Opus 5.
claude-opus-5,legal_agent_benchmark_public,23.58,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""bash and Python""}",False,model_card,verified,"Legal Agent Benchmark / public all-pass: 1,235 tested tasks after 16 exclusions; n=5; source reports +/-0.48."
claude-opus-5,legal_agent_benchmark_harvey_held_out,11.7,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max""}",False,model_card,verified,Legal Agent Benchmark / Harvey held-out all-pass: Proprietary Harvey held-out evaluation.
claude-opus-5,gdpval_aa_elo,1861.0,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""shell and web browsing""}",False,model_card,verified,"GDPval-AA v2 narrative / Claude Opus 5 / max: 220 tasks, 44 occupations, 9 industries; Artificial Analysis."
claude-opus-5,aa_briefcase_elo,1606.0,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""high""}",True,model_card,verified,AA-Briefcase narrative / Claude Opus 5 / high: Multi-week professional projects; Artificial Analysis.
claude-opus-5,toolathlon_verified,80.6,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""600+ tools across 32 applications""}",False,model_card,verified,Table 8.13.6.A / Claude Opus 5 / Pass@1: 108 tasks; three trials; internal table excludes null attempts. Opus 4.8 and Sonnet 5 internal values are candidates because official published leaderboard values include null attempts.
claude-sonnet-5,toolathlon_verified,71.6,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""600+ tools across 32 applications""}",False,model_card,verified,Toolathlon published leaderboard narrative / Claude Sonnet 5: Official published leaderboard including null attempts.
claude-opus-5,automation_bench_private_heldout,26.0,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max""}",False,model_card,verified,Figure 8.13.7.A / Claude Opus 5 / max: Private held-out set across 47 applications; deterministic assertions.
claude-opus-5,arc_agi_1,97.5,https://arcprize.org/leaderboard,"{""effort"":""max""}",False,leaderboard,verified,Figure 8.14 narrative / ARC-AGI-1 / Claude Opus 5: Verified semi-private ARC Prize score.
claude-opus-5,arc_agi_2,90.42,https://arcprize.org/leaderboard,"{""effort"":""max""}",False,leaderboard,verified,Figure 8.14 narrative / ARC-AGI-2 / Claude Opus 5: Verified semi-private ARC Prize score.
claude-mythos,healthbench,59.6,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""none""}",False,model_card,verified,Figure 8.15 / HealthBench / Claude Mythos Preview: Five trials; Opus 4.8 grader; no customized system prompt.
claude-opus-4.8,healthbench,58.8,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""none""}",False,model_card,verified,Figure 8.15 / HealthBench / Claude Opus 4.8: Five trials; Opus 4.8 grader; no customized system prompt.
claude-sonnet-5,healthbench,59.2,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""none""}",False,model_card,verified,Figure 8.15 / HealthBench / Claude Sonnet 5: Five trials; Opus 4.8 grader; no customized system prompt.
claude-opus-5,healthbench,67.1,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""none""}",False,model_card,verified,Figure 8.15 / HealthBench / Claude Opus 5: Five trials; Opus 4.8 grader; no customized system prompt.
claude-opus-5,healthbench_length_adjusted,57.8,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""none""}",False,model_card,verified,Figure 8.15 / HealthBench Length-Adjusted / Claude Opus 5: Five trials; Opus 4.8 grader; no customized system prompt.
claude-mythos,healthbench_professional,67.2,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""none""}",False,model_card,verified,Figure 8.15 / HealthBench Professional / Claude Mythos Preview: Five trials; Opus 4.8 grader; no customized system prompt.
claude-opus-4.8,healthbench_professional,60.3,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""none""}",False,model_card,verified,Figure 8.15 / HealthBench Professional / Claude Opus 4.8: Five trials; Opus 4.8 grader; no customized system prompt.
claude-sonnet-5,healthbench_professional,62.4,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""none""}",False,model_card,verified,Figure 8.15 / HealthBench Professional / Claude Sonnet 5: Five trials; Opus 4.8 grader; no customized system prompt.
claude-opus-5,healthbench_professional,73.4,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""none""}",False,model_card,verified,Figure 8.15 / HealthBench Professional / Claude Opus 5: Five trials; Opus 4.8 grader; no customized system prompt.
claude-mythos,gmmlu,92.9,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""none""}",False,model_card,verified,Figure 8.16 / GMMLU / Claude Mythos Preview: 42 languages; one trial.
claude-opus-4.8,gmmlu,90.9,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""none""}",False,model_card,verified,Figure 8.16 / GMMLU / Claude Opus 4.8: 42 languages; one trial.
claude-sonnet-5,gmmlu,89.0,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""none""}",False,model_card,verified,Figure 8.16 / GMMLU / Claude Sonnet 5: 42 languages; one trial.
claude-opus-5,gmmlu,92.5,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""none""}",False,model_card,verified,Figure 8.16 / GMMLU / Claude Opus 5: 42 languages; one trial.
claude-opus-5,milu,92.1,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""none""}",False,model_card,verified,Figure 8.16 / MILU / Claude Opus 5: 11 languages; five trials.
claude-opus-5,include_base_44,89.8,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""none""}",False,model_card,verified,Figure 8.16 / INCLUDE-base-44 / Claude Opus 5: 44 languages; five trials.
claude-opus-4.8,biomysterybench_human_solvable_revised_2026_07,0.885,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""benchmark-specific""}",False,model_card,verified,Figure 8.17.6.A / BioMysteryBench Human Solvable Revised / Claude Opus 4.8: Revised subset after 3 removals. Source label 88.5%; normalized to 0.885 for the 0-1 score metric.
claude-sonnet-5,biomysterybench_human_solvable_revised_2026_07,0.875,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""benchmark-specific""}",False,model_card,verified,Figure 8.17.6.A / BioMysteryBench Human Solvable Revised / Claude Sonnet 5: Revised subset after 3 removals. Source label 87.5%; normalized to 0.875 for the 0-1 score metric.
claude-opus-5,biomysterybench_human_solvable_revised_2026_07,0.901,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""benchmark-specific""}",False,model_card,verified,Figure 8.17.6.A / BioMysteryBench Human Solvable Revised / Claude Opus 5: Revised subset after 3 removals. Source label 90.1%; normalized to 0.901 for the 0-1 score metric.
claude-opus-4.8,biomysterybench_human_difficult_revised_2026_07,0.424,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""benchmark-specific""}",False,model_card,verified,Figure 8.17.6.A / BioMysteryBench Human Difficult Revised / Claude Opus 4.8: Revised subset after 6 removals. Source label 42.4%; normalized to 0.424 for the 0-1 score metric.
claude-sonnet-5,biomysterybench_human_difficult_revised_2026_07,0.341,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""benchmark-specific""}",False,model_card,verified,Figure 8.17.6.A / BioMysteryBench Human Difficult Revised / Claude Sonnet 5: Revised subset after 6 removals. Source label 34.1%; normalized to 0.341 for the 0-1 score metric.
claude-opus-5,biomysterybench_human_difficult_revised_2026_07,0.494,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""benchmark-specific""}",False,model_card,verified,Figure 8.17.6.A / BioMysteryBench Human Difficult Revised / Claude Opus 5: Revised subset after 6 removals. Source label 49.4%; normalized to 0.494 for the 0-1 score metric.
claude-opus-5,spatialbench_verified,0.725,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""benchmark-specific""}",False,model_card,verified,Figure 8.17.6.A / SpatialBench Verified / Claude Opus 5: 115 problems. Source label 72.5%; normalized to 0.725 for the 0-1 score metric.
claude-opus-5,singlecellbench,0.606,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""benchmark-specific""}",False,model_card,verified,Figure 8.17.6.A / SingleCellBench / Claude Opus 5: 195 problems. Source label 60.6%; normalized to 0.606 for the 0-1 score metric.
claude-opus-5,proteingym_hard,0.477,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""benchmark-specific""}",False,model_card,verified,Figure 8.17.6.A / ProteinGym Hard / Claude Opus 5: Rank correlation. Source label 47.7%; normalized to 0.477 for the 0-1 score metric.
claude-opus-4.8,protein_design_internal,0.32,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""benchmark-specific""}",False,model_card,verified,"Figure 8.17.6.A / Protein Design / Claude Opus 4.8: Combined constraint-satisfaction, folding-confidence, and novelty score. Source label 32.0%; normalized to 0.32 for the 0-1 score metric."
claude-sonnet-5,protein_design_internal,0.212,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""benchmark-specific""}",False,model_card,verified,"Figure 8.17.6.A / Protein Design / Claude Sonnet 5: Combined constraint-satisfaction, folding-confidence, and novelty score. Source label 21.2%; normalized to 0.212 for the 0-1 score metric."
claude-opus-5,protein_design_internal,0.425,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""benchmark-specific""}",False,model_card,verified,"Figure 8.17.6.A / Protein Design / Claude Opus 5: Combined constraint-satisfaction, folding-confidence, and novelty score. Source label 42.5%; normalized to 0.425 for the 0-1 score metric."
claude-opus-4.8,organic_chemistry_v2_internal,0.504,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""benchmark-specific""}",False,model_card,verified,"Figure 8.17.6.A / Organic Chemistry V2 / Claude Opus 4.8: Opus 5 timed out on 10/1,260 attempts; excluded; stated effect <= +/-0.006. Source label 50.4%; normalized to 0.504 for the 0-1 score metric."
claude-sonnet-5,organic_chemistry_v2_internal,0.406,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""benchmark-specific""}",False,model_card,verified,"Figure 8.17.6.A / Organic Chemistry V2 / Claude Sonnet 5: Opus 5 timed out on 10/1,260 attempts; excluded; stated effect <= +/-0.006. Source label 40.6%; normalized to 0.406 for the 0-1 score metric."
claude-opus-5,organic_chemistry_v2_internal,0.616,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""benchmark-specific""}",False,model_card,verified,"Figure 8.17.6.A / Organic Chemistry V2 / Claude Opus 5: Opus 5 timed out on 10/1,260 attempts; excluded; stated effect <= +/-0.006. Source label 61.6%; normalized to 0.616 for the 0-1 score metric."
claude-opus-5,protocol_troubleshooting_internal,0.611,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""benchmark-specific""}",False,model_card,verified,"Figure 8.17.6.A / Protocol Troubleshooting / Claude Opus 5: Bash, file editor, and web search. Source label 61.1%; normalized to 0.611 for the 0-1 score metric."
claude-opus-4.8,protocol_understanding_internal,0.623,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""benchmark-specific""}",False,model_card,verified,Figure 8.17.6.A / Protocol Understanding / Claude Opus 4.8: Benchling protocol-understanding variant. Source label 62.3%; normalized to 0.623 for the 0-1 score metric.
claude-sonnet-5,protocol_understanding_internal,0.605,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""benchmark-specific""}",False,model_card,verified,Figure 8.17.6.A / Protocol Understanding / Claude Sonnet 5: Benchling protocol-understanding variant. Source label 60.5%; normalized to 0.605 for the 0-1 score metric.
claude-opus-5,protocol_understanding_internal,0.784,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""tools"":""benchmark-specific""}",False,model_card,verified,Figure 8.17.6.A / Protocol Understanding / Claude Opus 5: Benchling protocol-understanding variant. Source label 78.4%; normalized to 0.784 for the 0-1 score metric.
claude-opus-5,long_form_virology_task_1_sequence_design,0.89,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""source does not state""}",False,model_card,verified,Figure 2.2.4.A / Long-form virology / Task 1 / Sequence design / Claude Opus 5: Two long-form virology tasks.
claude-opus-5,long_form_virology_task_1_protocol_design,0.91,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""source does not state""}",False,model_card,verified,Figure 2.2.4.A / Long-form virology / Task 1 / Protocol design / Claude Opus 5: Two long-form virology tasks.
claude-opus-5,long_form_virology_task_1_end_to_end,0.802,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""source does not state""}",False,model_card,verified,Figure 2.2.4.A / Long-form virology / Task 1 / End-to-end / Claude Opus 5: Two long-form virology tasks.
claude-opus-5,long_form_virology_task_2_sequence_design,1.0,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""source does not state""}",False,model_card,verified,Figure 2.2.4.A / Long-form virology / Task 2 / Sequence design / Claude Opus 5: Two long-form virology tasks.
claude-opus-5,long_form_virology_task_2_protocol_design,0.87,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""source does not state""}",False,model_card,verified,Figure 2.2.4.A / Long-form virology / Task 2 / Protocol design / Claude Opus 5: Two long-form virology tasks.
claude-opus-5,long_form_virology_task_2_end_to_end,0.872,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""source does not state""}",False,model_card,verified,Figure 2.2.4.A / Long-form virology / Task 2 / End-to-end / Claude Opus 5: Two long-form virology tasks. Task 2 Opus 5 narrative reports 0.872; the rendered figure rounds it to 0.87.
claude-opus-5,ai_rd_kernel_best_speedup,449.46,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""source does not state""}",False,model_card,verified,Table 2.3.5.A / Kernel best speedup / Claude Opus 5: Anthropic internal AI R&D task-based evaluation.
claude-opus-5,ai_rd_time_series_forecasting_mse,5.68,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""source does not state""}",False,model_card,verified,Table 2.3.5.A / Time-Series Forecasting MSE / Claude Opus 5: Anthropic internal AI R&D task-based evaluation.
claude-opus-5,ai_rd_llm_training_speedup,68.54,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""source does not state""}",False,model_card,verified,Table 2.3.5.A / LLM Training easy speedup / Claude Opus 5: Anthropic internal AI R&D task-based evaluation.
claude-opus-5,ai_rd_llm_training_hard_speedup,14.19,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""source does not state""}",False,model_card,verified,Table 2.3.5.A / LLM Training hard speedup / Claude Opus 5: Anthropic internal AI R&D task-based evaluation.
claude-opus-5,ai_rd_quadruped_rl,31.3,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""source does not state""}",False,model_card,verified,Table 2.3.5.A / Quadruped RL / Claude Opus 5: Anthropic internal AI R&D task-based evaluation.
claude-opus-5,ai_rd_novel_compiler,80.91,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""source does not state""}",False,model_card,verified,Table 2.3.5.A / Novel Compiler complex-test pass rate / Claude Opus 5: Anthropic internal AI R&D task-based evaluation.
claude-opus-5,exploitbench,70.0,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""source does not state""}",False,model_card,verified,Figure 3.3.1.A / ExploitBench Cap% / Claude Opus 5: 41 V8 vulnerabilities; five trials; Cap% uses a random three-trial subset; safeguards off.
claude-opus-5,oss_fuzz_any_progress,79.4,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""source does not state""}",False,model_card,verified,Figure 3.3.2.A / OSS-Fuzz any progress / Claude Opus 5: Approximately 830 entry points from 228 projects; safeguards off.
claude-opus-5,oss_fuzz_control_flow_hijack_count,4.0,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""source does not state""}",False,model_card,verified,Figure 3.3.2.A / OSS-Fuzz control-flow hijack / Claude Opus 5: Count reaching grade 1.0 across approximately 830 entry points.
claude-opus-5,firefox_147_exploit_development_working_exploit,52.4,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""source does not state""}",False,model_card,verified,"Figure 3.3.3.A / Firefox 147 working exploit / Claude Opus 5: 50 crash categories, five trials each, 250 trials total; mitigations off."
claude-opus-5,firefox_147_exploit_development_any_success,87.2,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""source does not state""}",False,model_card,verified,Figure 3.3.3.A / Firefox 147 any success / Claude Opus 5: Grade >=0.5 across 250 trials; mitigations off.
claude-opus-4.8,exploitgym,13.808976,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""harness"":""6h wall-clock budget""}",False,model_card,verified,Figure 3.3.5.A / ExploitGym / Claude Opus 4.8 / 6h: 120/869 successful exploits; mitigations off; agent-judge verification. Exact percentage derived deterministically from printed count 120/869.
claude-opus-5,exploitgym,21.979287,https://www.anthropic.com/claude-opus-5-system-card,"{""effort"":""max"",""harness"":""6h wall-clock budget""}",False,model_card,verified,Figure 3.3.5.A / ExploitGym / Claude Opus 5 / 6h: 191/869 successful exploits; mitigations off; agent-judge verification. Exact percentage derived deterministically from printed count 191/869.
claude-fable-5,cursorbench_3_2,70.5,https://cursor.com/cursorbench,"{""effort"":""max"",""harness"":""Cursor agent""}",True,leaderboard,verified,"Official CursorBench table / Fable 5 Max: Official score; average cost/task $17.32; 103,525 tokens/task; 72 steps/task."
gpt-5.6-sol,cursorbench_3_2,67.2,https://cursor.com/cursorbench,"{""effort"":""max"",""harness"":""Cursor agent""}",True,leaderboard,verified,"Official CursorBench table / GPT-5.6 Sol Max: Official score; average cost/task $5.69; 28,320 tokens/task; 48 steps/task."
claude-opus-5,cursorbench_3_2,66.7,https://cursor.com/cursorbench,"{""effort"":""high"",""harness"":""Cursor agent""}",True,leaderboard,verified,"Official CursorBench table / Opus 5 High: Official score; average cost/task $3.91; 27,932 tokens/task; 48 steps/task."
gpt-5.6-terra,cursorbench_3_2,64.9,https://cursor.com/cursorbench,"{""effort"":""max"",""harness"":""Cursor agent""}",True,leaderboard,verified,"Official CursorBench table / GPT-5.6 Terra Max: Official score; average cost/task $2.31; 32,969 tokens/task; 47 steps/task."
gpt-5.6-luna,cursorbench_3_2,61.1,https://cursor.com/cursorbench,"{""effort"":""max"",""harness"":""Cursor agent""}",True,leaderboard,verified,"Official CursorBench table / GPT-5.6 Luna Max: Official score; average cost/task $0.39; 87,973 tokens/task; 61 steps/task."
kimi-k3,cursorbench_3_2,60.8,https://cursor.com/cursorbench,"{""effort"":""max"",""harness"":""Cursor agent""}",True,leaderboard,verified,"Official CursorBench table / Kimi K3 Max: Official score; average cost/task $2.7; 38,428 tokens/task; 57 steps/task."
gpt-5.5,cursorbench_3_2,58.4,https://cursor.com/cursorbench,"{""effort"":""xhigh"",""harness"":""Cursor agent""}",True,leaderboard,verified,"Official CursorBench table / GPT-5.5 Extra High: Official score; average cost/task $2.85; 17,534 tokens/task; 32 steps/task."
claude-opus-4.8,cursorbench_3_2,58.0,https://cursor.com/cursorbench,"{""effort"":""high"",""harness"":""Cursor agent""}",True,leaderboard,verified,"Official CursorBench table / Opus 4.8 High: Official score; average cost/task $3.15; 33,548 tokens/task; 33 steps/task."
claude-sonnet-5,cursorbench_3_2,56.9,https://cursor.com/cursorbench,"{""effort"":""high"",""harness"":""Cursor agent""}",True,leaderboard,verified,"Official CursorBench table / Sonnet 5 High: Official score; average cost/task $2.13; 39,483 tokens/task; 57 steps/task."
composer-2.5,cursorbench_3_2,56.1,https://cursor.com/cursorbench,"{""harness"":""Cursor agent""}",False,leaderboard,verified,"Official CursorBench table / Composer 2.5: Official score; average cost/task $0.44; 14,286 tokens/task; 33 steps/task."
glm-5.2,cursorbench_3_2,55.0,https://cursor.com/cursorbench,"{""effort"":""max"",""harness"":""Cursor agent""}",True,leaderboard,verified,"Official CursorBench table / GLM 5.2 Max: Official score; average cost/task $1.76; 35,946 tokens/task; 58 steps/task."
gemini-3.6-flash,cursorbench_3_2,51.2,https://cursor.com/cursorbench,"{""effort"":""medium"",""harness"":""Cursor agent""}",True,leaderboard,verified,"Official CursorBench table / Gemini 3.6 Flash Medium: Official score; average cost/task $1.48; 28,511 tokens/task; 62 steps/task."
kimi-k2.7-code,cursorbench_3_2,49.7,https://cursor.com/cursorbench,"{""harness"":""Cursor agent""}",False,leaderboard,verified,"Official CursorBench table / Kimi K2.7 Code: Official score; average cost/task $1.43; 31,247 tokens/task; 58 steps/task."
claude-opus-4.6,arc_agi_3,0.51,https://arcprize.org/leaderboard,"{""effort"":""max""}",True,leaderboard,verified,ARC Prize leaderboard / Anthropic Opus 4.6 (Max): Relative Human Action Efficiency; total cost $8866.2.
grok-4.20,arc_agi_3,0.09,https://arcprize.org/leaderboard,{},False,leaderboard,verified,ARC Prize leaderboard / Grok 4.20 (Beta Reasoning): Relative Human Action Efficiency; total cost $3775.
gpt-5.4,arc_agi_3,0.21,https://arcprize.org/leaderboard,"{""effort"":""high""}",False,leaderboard,verified,ARC Prize leaderboard / GPT-5.4 (High): Relative Human Action Efficiency; total cost $5187.41.
claude-opus-4.7,arc_agi_3,0.18,https://arcprize.org/leaderboard,"{""effort"":""high""}",False,leaderboard,verified,ARC Prize leaderboard / Opus 4.7 (High): Relative Human Action Efficiency; total cost $10000.
grok-4.5,arc_agi_3,0.3,https://arcprize.org/leaderboard,"{""effort"":""high""}",True,leaderboard,verified,ARC Prize leaderboard / Grok 4.5 (High): Relative Human Action Efficiency; total cost $6892.87.
claude-opus-5,arc_agi_3,30.16,https://arcprize.org/leaderboard,"{""effort"":""high""}",True,leaderboard,verified,ARC Prize leaderboard / Claude Opus 5 (High): Relative Human Action Efficiency; total cost $20657.4.
composer-2.5,terminal_bench,69.3,https://cursor.com/blog/composer-2-5,"{""effort"":""source does not state"",""harness"":""Cursor"",""sampling"":""pass@1"",""tools"":""agentic""}",False,official_blog,verified,Composer 2.5 release benchmark image / Terminal-Bench 2.0 / Composer 2.5: Cursor release benchmark image. Opus 4.7 and GPT-5.5 public evaluation values are self-reported.
composer-2,terminal_bench,61.7,https://cursor.com/blog/composer-2-5,"{""effort"":""source does not state"",""harness"":""Cursor"",""sampling"":""pass@1"",""tools"":""agentic""}",False,official_blog,verified,Composer 2.5 release benchmark image / Terminal-Bench 2.0 / Composer 2: Cursor release benchmark image. Opus 4.7 and GPT-5.5 public evaluation values are self-reported.
composer-2.5,swe_bench_multilingual,79.8,https://cursor.com/blog/composer-2-5,"{""effort"":""source does not state"",""harness"":""Cursor"",""sampling"":""pass@1"",""tools"":""agentic""}",False,official_blog,verified,Composer 2.5 release benchmark image / SWE-Bench Multilingual / Composer 2.5: Cursor release benchmark image. Opus 4.7 and GPT-5.5 public evaluation values are self-reported.
composer-2,swe_bench_multilingual,73.7,https://cursor.com/blog/composer-2-5,"{""effort"":""source does not state"",""harness"":""Cursor"",""sampling"":""pass@1"",""tools"":""agentic""}",False,official_blog,verified,Composer 2.5 release benchmark image / SWE-Bench Multilingual / Composer 2: Cursor release benchmark image. Opus 4.7 and GPT-5.5 public evaluation values are self-reported.
composer-2.5,cursorbench_3_1,63.2,https://cursor.com/blog/composer-2-5,"{""effort"":""source does not state"",""harness"":""Cursor"",""sampling"":""pass@1"",""tools"":""agentic""}",False,official_blog,verified,Composer 2.5 release benchmark image / CursorBench 3.1 / Composer 2.5: Cursor release benchmark image. Opus 4.7 and GPT-5.5 public evaluation values are self-reported.
composer-2,cursorbench_3_1,52.2,https://cursor.com/blog/composer-2-5,"{""effort"":""source does not state"",""harness"":""Cursor"",""sampling"":""pass@1"",""tools"":""agentic""}",False,official_blog,verified,Composer 2.5 release benchmark image / CursorBench 3.1 / Composer 2: Cursor release benchmark image. Opus 4.7 and GPT-5.5 public evaluation values are self-reported.
grok-4.6,cursorbench_3_2,69.9,https://x.ai/news/grok-4-6,"{""context"":""unknown"",""effort"":""high"",""harness"":""Cursor first-party agent harness"",""judge"":""CursorBench task score"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""Cursor coding-agent tools""}",True,official_blog,verified,release46 CursorBench v3.2 Score (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
gemini-3.7-flash,cursorbench_3_2,59.0,https://cursor.com/cursorbench,"{""effort"":""medium"",""harness"":""Cursor agent""}",True,leaderboard,verified,"CursorBench 3.2 current leaderboard / rank 28 / Gemini 3.7 Flash Medium: Average cost/task $ 0.95; tokens/task 30,953; steps/task 82."
composer-2,cursorbench_3_0,61.3,https://cursor.com/resources/Composer2.pdf,"{""effort"":""source does not state"",""harness"":""Cursor"",""sampling"":""pass@1"",""tools"":""Cursor coding agent""}",False,tech_report,verified,"Composer 2 technical report Table 1 / CursorBench 3.0 / Composer 2 / value 1: For third-party public benchmarks, slash-separated values are Cursor/official-harness then self-reported. SWE prompts prepend 'please solve this github issue'; Terminal prompts add solution-format instructions."
composer-1.5,cursorbench_3_0,44.2,https://cursor.com/resources/Composer2.pdf,"{""effort"":""source does not state"",""harness"":""Cursor"",""sampling"":""pass@1"",""tools"":""Cursor coding agent""}",False,tech_report,verified,"Composer 2 technical report Table 1 / CursorBench 3.0 / Composer 1.5 / value 1: For third-party public benchmarks, slash-separated values are Cursor/official-harness then self-reported. SWE prompts prepend 'please solve this github issue'; Terminal prompts add solution-format instructions."
composer-1.5,swe_bench_multilingual,65.9,https://cursor.com/resources/Composer2.pdf,"{""effort"":""source does not state"",""harness"":""Cursor harness"",""sampling"":""pass@1"",""tools"":""Cursor coding agent""}",False,tech_report,verified,"Composer 2 technical report Table 1 / SWE-Bench Multilingual / Composer 1.5 / value 1: For third-party public benchmarks, slash-separated values are Cursor/official-harness then self-reported. SWE prompts prepend 'please solve this github issue'; Terminal prompts add solution-format instructions."
composer-1.5,terminal_bench,47.9,https://cursor.com/resources/Composer2.pdf,"{""effort"":""source does not state"",""harness"":""Harbor / Cursor-selected official value"",""sampling"":""pass@1"",""tools"":""Cursor coding agent""}",False,tech_report,verified,"Composer 2 technical report Table 1 / Terminal-Bench 2.0 / Composer 1.5 / value 1: For third-party public benchmarks, slash-separated values are Cursor/official-harness then self-reported. SWE prompts prepend 'please solve this github issue'; Terminal prompts add solution-format instructions."
composer-1,cursorbench_3_0,38.0,https://cursor.com/resources/Composer2.pdf,"{""effort"":""source does not state"",""harness"":""Cursor"",""sampling"":""pass@1"",""tools"":""Cursor coding agent""}",False,tech_report,verified,"Composer 2 technical report Table 1 / CursorBench 3.0 / Composer 1 / value 1: For third-party public benchmarks, slash-separated values are Cursor/official-harness then self-reported. SWE prompts prepend 'please solve this github issue'; Terminal prompts add solution-format instructions."
composer-1,swe_bench_multilingual,56.9,https://cursor.com/resources/Composer2.pdf,"{""effort"":""source does not state"",""harness"":""Cursor harness"",""sampling"":""pass@1"",""tools"":""Cursor coding agent""}",False,tech_report,verified,"Composer 2 technical report Table 1 / SWE-Bench Multilingual / Composer 1 / value 1: For third-party public benchmarks, slash-separated values are Cursor/official-harness then self-reported. SWE prompts prepend 'please solve this github issue'; Terminal prompts add solution-format instructions."
composer-1,terminal_bench,40.0,https://cursor.com/resources/Composer2.pdf,"{""effort"":""source does not state"",""harness"":""Harbor / Cursor-selected official value"",""sampling"":""pass@1"",""tools"":""Cursor coding agent""}",False,tech_report,verified,"Composer 2 technical report Table 1 / Terminal-Bench 2.0 / Composer 1 / value 1: For third-party public benchmarks, slash-separated values are Cursor/official-harness then self-reported. SWE prompts prepend 'please solve this github issue'; Terminal prompts add solution-format instructions."
claude-opus-4.6,cursorbench_3_0,58.2,https://cursor.com/resources/Composer2.pdf,"{""effort"":""high"",""harness"":""Cursor"",""sampling"":""pass@1"",""tools"":""Cursor coding agent""}",False,tech_report,verified,"Composer 2 technical report Table 1 / CursorBench 3.0 / Opus 4.6 High / value 1: For third-party public benchmarks, slash-separated values are Cursor/official-harness then self-reported. SWE prompts prepend 'please solve this github issue'; Terminal prompts add solution-format instructions."
claude-opus-4.5,cursorbench_3_0,48.4,https://cursor.com/resources/Composer2.pdf,"{""effort"":""high"",""harness"":""Cursor"",""sampling"":""pass@1"",""tools"":""Cursor coding agent""}",True,tech_report,verified,"Composer 2 technical report Table 1 / CursorBench 3.0 / Opus 4.5 High / value 1: For third-party public benchmarks, slash-separated values are Cursor/official-harness then self-reported. SWE prompts prepend 'please solve this github issue'; Terminal prompts add solution-format instructions."
gpt-5.4,cursorbench_3_0,63.9,https://cursor.com/resources/Composer2.pdf,"{""effort"":""high"",""harness"":""Cursor"",""sampling"":""pass@1"",""tools"":""Cursor coding agent""}",False,tech_report,verified,"Composer 2 technical report Table 1 / CursorBench 3.0 / GPT-5.4 / value 1: For third-party public benchmarks, slash-separated values are Cursor/official-harness then self-reported. SWE prompts prepend 'please solve this github issue'; Terminal prompts add solution-format instructions."
gpt-5.4,swe_bench_multilingual,76.8,https://cursor.com/resources/Composer2.pdf,"{""effort"":""high"",""harness"":""Cursor harness"",""sampling"":""pass@1"",""tools"":""Cursor coding agent""}",False,tech_report,verified,"Composer 2 technical report Table 1 / SWE-Bench Multilingual / GPT-5.4 / value 1: For third-party public benchmarks, slash-separated values are Cursor/official-harness then self-reported. SWE prompts prepend 'please solve this github issue'; Terminal prompts add solution-format instructions."
gpt-5.3-codex,cursorbench_3_0,59.1,https://cursor.com/resources/Composer2.pdf,"{""effort"":""high"",""harness"":""Cursor"",""sampling"":""pass@1"",""tools"":""Cursor coding agent""}",False,tech_report,verified,"Composer 2 technical report Table 1 / CursorBench 3.0 / GPT-5.3 Codex / value 1: For third-party public benchmarks, slash-separated values are Cursor/official-harness then self-reported. SWE prompts prepend 'please solve this github issue'; Terminal prompts add solution-format instructions."
gpt-5.3-codex,swe_bench_multilingual,74.8,https://cursor.com/resources/Composer2.pdf,"{""effort"":""high"",""harness"":""Cursor harness"",""sampling"":""pass@1"",""tools"":""Cursor coding agent""}",False,tech_report,verified,"Composer 2 technical report Table 1 / SWE-Bench Multilingual / GPT-5.3 Codex / value 1: For third-party public benchmarks, slash-separated values are Cursor/official-harness then self-reported. SWE prompts prepend 'please solve this github issue'; Terminal prompts add solution-format instructions."
gpt-5.2,cursorbench_3_0,56.5,https://cursor.com/resources/Composer2.pdf,"{""effort"":""high"",""harness"":""Cursor"",""sampling"":""pass@1"",""tools"":""Cursor coding agent""}",False,tech_report,verified,"Composer 2 technical report Table 1 / CursorBench 3.0 / GPT-5.2 / value 1: For third-party public benchmarks, slash-separated values are Cursor/official-harness then self-reported. SWE prompts prepend 'please solve this github issue'; Terminal prompts add solution-format instructions."
glm-5,cursorbench_3_0,42.7,https://cursor.com/resources/Composer2.pdf,"{""effort"":""source does not state"",""harness"":""Cursor"",""sampling"":""pass@1"",""tools"":""Cursor coding agent""}",False,tech_report,verified,"Composer 2 technical report Table 1 / CursorBench 3.0 / GLM-5 / value 1: For third-party public benchmarks, slash-separated values are Cursor/official-harness then self-reported. SWE prompts prepend 'please solve this github issue'; Terminal prompts add solution-format instructions."
kimi-k2.5,cursorbench_3_0,36.0,https://cursor.com/resources/Composer2.pdf,"{""effort"":""source does not state"",""harness"":""Cursor"",""sampling"":""pass@1"",""tools"":""Cursor coding agent""}",False,tech_report,verified,"Composer 2 technical report Table 1 / CursorBench 3.0 / Kimi K2.5 / value 1: For third-party public benchmarks, slash-separated values are Cursor/official-harness then self-reported. SWE prompts prepend 'please solve this github issue'; Terminal prompts add solution-format instructions."
muse-spark-1.1,wmdp_bio_accuracy,89.0,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""source-specific"",""effort"":""high"",""harness"":""Meta safety/preparedness evaluation"",""mode"":""thinking"",""notes"":""Capability scorecard; Muse evaluated without system prompt; high reasoning."",""sampling"":""source-specific"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""benchmark-specified""}",False,tech_report,verified,Table 1 · p6; metric=accuracy_percent. Claude cyber-benchmark cells omitted where refusal compromised capability measurement.
muse-spark,wmdp_bio_accuracy,88.4,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""source-specific"",""effort"":""high"",""harness"":""Meta safety/preparedness evaluation"",""mode"":""thinking"",""notes"":""Capability scorecard; Muse evaluated without system prompt; high reasoning."",""sampling"":""source-specific"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""benchmark-specified""}",False,tech_report,verified,Table 1 · p6; metric=accuracy_percent. Claude cyber-benchmark cells omitted where refusal compromised capability measurement.
gpt-5.5,wmdp_bio_accuracy,90.4,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""source-specific"",""effort"":""xhigh"",""harness"":""Meta safety/preparedness evaluation"",""mode"":""thinking"",""notes"":""Capability scorecard; Muse evaluated without system prompt; high reasoning."",""sampling"":""source-specific"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""benchmark-specified""}",True,tech_report,verified,Table 1 · p6; metric=accuracy_percent. Claude cyber-benchmark cells omitted where refusal compromised capability measurement.
gemini-3.1-pro,wmdp_bio_accuracy,89.5,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""source-specific"",""effort"":""high"",""harness"":""Meta safety/preparedness evaluation"",""mode"":""thinking"",""notes"":""Capability scorecard; Muse evaluated without system prompt; high reasoning."",""sampling"":""source-specific"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""benchmark-specified""}",True,tech_report,verified,Table 1 · p6; metric=accuracy_percent. Claude cyber-benchmark cells omitted where refusal compromised capability measurement.
muse-spark-1.1,wmdp_chem_accuracy,87.0,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""source-specific"",""effort"":""high"",""harness"":""Meta safety/preparedness evaluation"",""mode"":""thinking"",""notes"":""Capability scorecard; Muse evaluated without system prompt; high reasoning."",""sampling"":""source-specific"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""benchmark-specified""}",False,tech_report,verified,Table 1 · p6; metric=accuracy_percent. Claude cyber-benchmark cells omitted where refusal compromised capability measurement.
muse-spark,wmdp_chem_accuracy,85.6,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""source-specific"",""effort"":""high"",""harness"":""Meta safety/preparedness evaluation"",""mode"":""thinking"",""notes"":""Capability scorecard; Muse evaluated without system prompt; high reasoning."",""sampling"":""source-specific"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""benchmark-specified""}",False,tech_report,verified,Table 1 · p6; metric=accuracy_percent. Claude cyber-benchmark cells omitted where refusal compromised capability measurement.
gpt-5.5,wmdp_chem_accuracy,85.5,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""source-specific"",""effort"":""xhigh"",""harness"":""Meta safety/preparedness evaluation"",""mode"":""thinking"",""notes"":""Capability scorecard; Muse evaluated without system prompt; high reasoning."",""sampling"":""source-specific"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""benchmark-specified""}",True,tech_report,verified,Table 1 · p6; metric=accuracy_percent. Claude cyber-benchmark cells omitted where refusal compromised capability measurement.
gemini-3.1-pro,wmdp_chem_accuracy,86.3,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""source-specific"",""effort"":""high"",""harness"":""Meta safety/preparedness evaluation"",""mode"":""thinking"",""notes"":""Capability scorecard; Muse evaluated without system prompt; high reasoning."",""sampling"":""source-specific"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""benchmark-specified""}",True,tech_report,verified,Table 1 · p6; metric=accuracy_percent. Claude cyber-benchmark cells omitted where refusal compromised capability measurement.
muse-spark-1.1,lab_bench_protocolqa_accuracy,88.0,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""source-specific"",""effort"":""high"",""harness"":""Meta safety/preparedness evaluation"",""mode"":""thinking"",""notes"":""Capability scorecard; Muse evaluated without system prompt; high reasoning."",""sampling"":""source-specific"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""benchmark-specified""}",False,tech_report,verified,Table 1 · p6; metric=accuracy_percent. Claude cyber-benchmark cells omitted where refusal compromised capability measurement.
muse-spark,lab_bench_protocolqa_accuracy,87.3,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""source-specific"",""effort"":""high"",""harness"":""Meta safety/preparedness evaluation"",""mode"":""thinking"",""notes"":""Capability scorecard; Muse evaluated without system prompt; high reasoning."",""sampling"":""source-specific"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""benchmark-specified""}",False,tech_report,verified,Table 1 · p6; metric=accuracy_percent. Claude cyber-benchmark cells omitted where refusal compromised capability measurement.
gpt-5.5,lab_bench_protocolqa_accuracy,78.2,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""source-specific"",""effort"":""xhigh"",""harness"":""Meta safety/preparedness evaluation"",""mode"":""thinking"",""notes"":""Capability scorecard; Muse evaluated without system prompt; high reasoning."",""sampling"":""source-specific"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""benchmark-specified""}",True,tech_report,verified,Table 1 · p6; metric=accuracy_percent. Claude cyber-benchmark cells omitted where refusal compromised capability measurement.
gemini-3.1-pro,lab_bench_protocolqa_accuracy,88.9,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""source-specific"",""effort"":""high"",""harness"":""Meta safety/preparedness evaluation"",""mode"":""thinking"",""notes"":""Capability scorecard; Muse evaluated without system prompt; high reasoning."",""sampling"":""source-specific"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""benchmark-specified""}",True,tech_report,verified,Table 1 · p6; metric=accuracy_percent. Claude cyber-benchmark cells omitted where refusal compromised capability measurement.
muse-spark-1.1,cybench,92.9,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""source-specific"",""effort"":""high"",""harness"":""Meta safety/preparedness evaluation"",""mode"":""thinking"",""notes"":""Capability scorecard; Muse evaluated without system prompt; high reasoning."",""sampling"":""pass@1"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""benchmark-specified""}",False,tech_report,verified,Table 1 · p6; metric=pass@1_percent. Claude cyber-benchmark cells omitted where refusal compromised capability measurement.
muse-spark,cybench,65.4,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""source-specific"",""effort"":""high"",""harness"":""Meta safety/preparedness evaluation"",""mode"":""thinking"",""notes"":""Capability scorecard; Muse evaluated without system prompt; high reasoning."",""sampling"":""pass@1"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""benchmark-specified""}",False,tech_report,verified,Table 1 · p6; metric=pass@1_percent. Claude cyber-benchmark cells omitted where refusal compromised capability measurement.
gpt-5.5,cybench,100.0,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""source-specific"",""effort"":""xhigh"",""harness"":""Meta safety/preparedness evaluation"",""mode"":""thinking"",""notes"":""Capability scorecard; Muse evaluated without system prompt; high reasoning."",""sampling"":""pass@1"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""benchmark-specified""}",False,tech_report,verified,Table 1 · p6; metric=pass@1_percent. Claude cyber-benchmark cells omitted where refusal compromised capability measurement.
claude-opus-4.8,cybench,95.0,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""source-specific"",""effort"":""max"",""harness"":""Meta safety/preparedness evaluation"",""mode"":""thinking"",""notes"":""Capability scorecard; Muse evaluated without system prompt; high reasoning."",""sampling"":""pass@1"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""benchmark-specified""}",False,tech_report,verified,Table 1 · p6; metric=pass@1_percent. Claude cyber-benchmark cells omitted where refusal compromised capability measurement.
muse-spark-1.1,cybergym,59.0,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""source-specific"",""effort"":""high"",""harness"":""Meta safety/preparedness evaluation"",""mode"":""thinking"",""notes"":""Capability scorecard; Muse evaluated without system prompt; high reasoning."",""sampling"":""pass@1"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""benchmark-specified""}",False,tech_report,verified,Table 1 · p6; metric=pass@1_percent. Claude cyber-benchmark cells omitted where refusal compromised capability measurement.
muse-spark,cybergym,43.5,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""source-specific"",""effort"":""high"",""harness"":""Meta safety/preparedness evaluation"",""mode"":""thinking"",""notes"":""Capability scorecard; Muse evaluated without system prompt; high reasoning."",""sampling"":""pass@1"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""benchmark-specified""}",False,tech_report,verified,Table 1 · p6; metric=pass@1_percent. Claude cyber-benchmark cells omitted where refusal compromised capability measurement.
muse-spark-1.1,exploitgym,0.8,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""source-specific"",""effort"":""high"",""harness"":""Meta safety/preparedness evaluation"",""mode"":""thinking"",""notes"":""Capability scorecard; Muse evaluated without system prompt; high reasoning."",""sampling"":""pass@1"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""benchmark-specified""}",False,tech_report,verified,Table 1 · p6; metric=pass@1_percent. Claude cyber-benchmark cells omitted where refusal compromised capability measurement.
gemini-3.1-pro,exploitgym,1.4,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""source-specific"",""effort"":""high"",""harness"":""Meta safety/preparedness evaluation"",""mode"":""thinking"",""notes"":""Capability scorecard; Muse evaluated without system prompt; high reasoning."",""sampling"":""pass@1"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""benchmark-specified""}",False,tech_report,verified,Table 1 · p6; metric=pass@1_percent. Claude cyber-benchmark cells omitted where refusal compromised capability measurement.
kimi-k2.5,wmdp_bio_accuracy,86.8,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""source-specific"",""effort"":""default"",""harness"":""Meta safety/preparedness evaluation"",""mode"":""thinking"",""notes"":""Exact bar labels."",""sampling"":""source-specific"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""benchmark-specified""}",True,tech_report,verified,Figure 2 · p13; metric=accuracy_percent.
kimi-k2.5,wmdp_chem_accuracy,82.0,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""source-specific"",""effort"":""default"",""harness"":""Meta safety/preparedness evaluation"",""mode"":""thinking"",""notes"":""Exact bar labels."",""sampling"":""source-specific"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""benchmark-specified""}",True,tech_report,verified,Figure 2 · p13; metric=accuracy_percent.
kimi-k2.5,lab_bench_protocolqa_accuracy,79.5,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""source-specific"",""effort"":""default"",""harness"":""Meta safety/preparedness evaluation"",""mode"":""thinking"",""notes"":""Muse Spark and Kimi include an abstention option; accuracy excludes abstentions."",""sampling"":""source-specific"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""benchmark-specified""}",False,tech_report,verified,Figure 3 · p14; metric=accuracy_percent. Asterisk marks abstention-option variants.
muse-spark-1.1,hle_tools,62.1,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""xhigh"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""general tools""}",True,tech_report,verified,Figure 44 · HLE w/tools; metric=headline_metric. Exact printed value in the general-capability summary.
muse-spark-1.1,hle,52.2,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""xhigh"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""none""}",True,tech_report,verified,Figure 44 · HLE no tools; metric=headline_metric. Exact printed value in the general-capability summary.
muse-spark-1.1,mrcr_v2_8needle_512k_1m,54.1,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""xhigh"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""none""}",True,tech_report,verified,"Figure 44 · MRCR v2 8-needle, 512K-1M; metric=headline_metric. Exact printed value in the general-capability summary."
gemini-3.1-pro,mrcr_v2_8needle_512k_1m,26.3,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""high"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""none""}",True,tech_report,verified,"Figure 44 · MRCR v2 8-needle, 512K-1M; metric=headline_metric. Exact printed value in the general-capability summary."
gpt-5.5,mrcr_v2_8needle_512k_1m,74.0,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""xhigh"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""none""}",True,tech_report,verified,"Figure 44 · MRCR v2 8-needle, 512K-1M; metric=headline_metric. Exact printed value in the general-capability summary."
muse-spark-1.1,mcpatlas_full_1000,88.1,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""xhigh"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""MCP tool servers""}",True,tech_report,verified,"Figure 44 · MCP-Atlas full 1,000 tasks; metric=headline_metric. Exact printed value in the general-capability summary."
muse-spark,mcpatlas_full_1000,82.2,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""default"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""MCP tool servers""}",True,tech_report,verified,"Figure 44 · MCP-Atlas full 1,000 tasks; metric=headline_metric. Exact printed value in the general-capability summary."
gemini-3.1-pro,mcpatlas_full_1000,78.2,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""high"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""MCP tool servers""}",True,tech_report,verified,"Figure 44 · MCP-Atlas full 1,000 tasks; metric=headline_metric. Exact printed value in the general-capability summary."
claude-opus-4.8,mcpatlas_full_1000,82.2,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""max"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""MCP tool servers""}",True,tech_report,verified,"Figure 44 · MCP-Atlas full 1,000 tasks; metric=headline_metric. Exact printed value in the general-capability summary."
gpt-5.5,mcpatlas_full_1000,75.3,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""xhigh"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""MCP tool servers""}",True,tech_report,verified,"Figure 44 · MCP-Atlas full 1,000 tasks; metric=headline_metric. Exact printed value in the general-capability summary."
muse-spark-1.1,toolathlon_verified,75.6,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""xhigh"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""benchmark tools""}",True,tech_report,verified,Figure 44 · Toolathlon-Verified; metric=headline_metric. Exact printed value in the general-capability summary.
muse-spark,toolathlon_verified,49.4,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""default"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""benchmark tools""}",True,tech_report,verified,Figure 44 · Toolathlon-Verified; metric=headline_metric. Exact printed value in the general-capability summary.
gemini-3.1-pro,toolathlon_verified,61.1,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""high"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""benchmark tools""}",True,tech_report,verified,Figure 44 · Toolathlon-Verified; metric=headline_metric. Exact printed value in the general-capability summary.
muse-spark-1.1,osworld_verified,80.8,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""xhigh"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""computer use""}",True,tech_report,verified,Figure 44 · OSWorld-Verified; metric=headline_metric. Exact printed value in the general-capability summary.
muse-spark,osworld_verified,53.3,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""default"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""computer use""}",True,tech_report,verified,Figure 44 · OSWorld-Verified; metric=headline_metric. Exact printed value in the general-capability summary.
gemini-3.1-pro,osworld_verified,76.2,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""high"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""computer use""}",True,tech_report,verified,Figure 44 · OSWorld-Verified; metric=headline_metric. Exact printed value in the general-capability summary.
muse-spark-1.1,osworld_2_0_binary,14.2,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""xhigh"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""computer use""}",True,tech_report,verified,Figure 44 · OSWorld 2.0 binary; metric=headline_metric. Exact printed value in the general-capability summary.
gemini-3.1-pro,osworld_2_0_binary,7.8,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""high"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""computer use""}",True,tech_report,verified,Figure 44 · OSWorld 2.0 binary; metric=headline_metric. Exact printed value in the general-capability summary.
claude-opus-4.8,osworld_2_0_binary,20.6,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""max"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""computer use""}",True,tech_report,verified,Figure 44 · OSWorld 2.0 binary; metric=headline_metric. Exact printed value in the general-capability summary.
gpt-5.5,osworld_2_0_binary,13.9,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""xhigh"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""computer use""}",True,tech_report,verified,Figure 44 · OSWorld 2.0 binary; metric=headline_metric. Exact printed value in the general-capability summary.
muse-spark-1.1,osworld_2_0,47.3,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""xhigh"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""computer use""}",True,tech_report,verified,Figure 44 · OSWorld 2.0 partial; metric=headline_metric. Exact printed value in the general-capability summary.
gemini-3.1-pro,osworld_2_0,30.6,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""high"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""computer use""}",True,tech_report,verified,Figure 44 · OSWorld 2.0 partial; metric=headline_metric. Exact printed value in the general-capability summary.
muse-spark-1.1,webarena_verified_full,69.0,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""xhigh"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""browser""}",True,tech_report,verified,Figure 44 · WebArena-Verified; metric=headline_metric. Exact printed value in the general-capability summary.
muse-spark,webarena_verified_full,59.0,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""default"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""browser""}",True,tech_report,verified,Figure 44 · WebArena-Verified; metric=headline_metric. Exact printed value in the general-capability summary.
gemini-3.1-pro,webarena_verified_full,69.0,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""high"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""browser""}",True,tech_report,verified,Figure 44 · WebArena-Verified; metric=headline_metric. Exact printed value in the general-capability summary.
claude-opus-4.8,webarena_verified_full,71.2,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""max"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""browser""}",True,tech_report,verified,Figure 44 · WebArena-Verified; metric=headline_metric. Exact printed value in the general-capability summary.
gpt-5.5,webarena_verified_full,67.0,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""xhigh"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""browser""}",True,tech_report,verified,Figure 44 · WebArena-Verified; metric=headline_metric. Exact printed value in the general-capability summary.
muse-spark-1.1,deepsearchqa_f1,84.9,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""xhigh"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""search/browser""}",True,tech_report,verified,Figure 44 · DeepSearchQA F1; metric=headline_metric. Exact printed value in the general-capability summary.
muse-spark-1.1,gdpval_aa_v2_elo,1381.0,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""xhigh"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""benchmark-specified""}",True,tech_report,verified,Figure 44 · GDPval-AA v2 Elo; metric=headline_metric. Exact printed value in the general-capability summary.
muse-spark,gdpval_aa_v2_elo,1145.0,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""default"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""benchmark-specified""}",True,tech_report,verified,Figure 44 · GDPval-AA v2 Elo; metric=headline_metric. Exact printed value in the general-capability summary.
gemini-3.1-pro,gdpval_aa_v2_elo,963.0,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""high"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""benchmark-specified""}",True,tech_report,verified,Figure 44 · GDPval-AA v2 Elo; metric=headline_metric. Exact printed value in the general-capability summary.
claude-opus-4.8,gdpval_aa_v2_elo,1600.0,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""max"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""benchmark-specified""}",True,tech_report,verified,Figure 44 · GDPval-AA v2 Elo; metric=headline_metric. Exact printed value in the general-capability summary.
gpt-5.5,gdpval_aa_v2_elo,1494.0,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""xhigh"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""benchmark-specified""}",True,tech_report,verified,Figure 44 · GDPval-AA v2 Elo; metric=headline_metric. Exact printed value in the general-capability summary.
muse-spark-1.1,job_bench,54.7,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""xhigh"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""benchmark-specified""}",True,tech_report,verified,Figure 44 · JobBench; metric=headline_metric. Exact printed value in the general-capability summary.
muse-spark,job_bench,17.0,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""default"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""benchmark-specified""}",True,tech_report,verified,Figure 44 · JobBench; metric=headline_metric. Exact printed value in the general-capability summary.
gemini-3.1-pro,job_bench,15.9,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""high"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""benchmark-specified""}",True,tech_report,verified,Figure 44 · JobBench; metric=headline_metric. Exact printed value in the general-capability summary.
muse-spark-1.1,finance_agent_v2,57.2,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""xhigh"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""finance-agent tools""}",True,tech_report,verified,Figure 44 · Finance Agent v2; metric=headline_metric. Exact printed value in the general-capability summary.
gemini-3.1-pro,finance_agent_v2,43.0,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""high"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""finance-agent tools""}",True,tech_report,verified,Figure 44 · Finance Agent v2; metric=headline_metric. Exact printed value in the general-capability summary.
muse-spark-1.1,terminal_bench_2_1,80.0,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""xhigh"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""terminal""}",True,tech_report,verified,Figure 44 · Terminal-Bench 2.1; metric=headline_metric. Exact printed value in the general-capability summary.
muse-spark,terminal_bench_2_1,67.3,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""default"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""terminal""}",True,tech_report,verified,Figure 44 · Terminal-Bench 2.1; metric=headline_metric. Exact printed value in the general-capability summary.
muse-spark-1.1,swe_bench_pro,61.5,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""xhigh"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""agentic coding""}",True,tech_report,verified,Figure 44 · SWE-Bench Pro; metric=headline_metric. Exact printed value in the general-capability summary.
muse-spark-1.1,deep_swe_v1_1,53.3,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""xhigh"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""agentic coding""}",True,tech_report,verified,Figure 44 · DeepSWE 1.1; metric=headline_metric. Exact printed value in the general-capability summary.
muse-spark,deep_swe_v1_1,10.0,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""default"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""agentic coding""}",True,tech_report,verified,Figure 44 · DeepSWE 1.1; metric=headline_metric. Exact printed value in the general-capability summary.
muse-spark-1.1,healthbench_professional_length_adjusted,59.3,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""xhigh"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""none""}",True,tech_report,verified,Figure 44 · HealthBench Professional length-adjusted; metric=headline_metric. Exact printed value in the general-capability summary.
muse-spark,healthbench_professional_length_adjusted,54.1,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""default"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""none""}",True,tech_report,verified,Figure 44 · HealthBench Professional length-adjusted; metric=headline_metric. Exact printed value in the general-capability summary.
gemini-3.1-pro,healthbench_professional_length_adjusted,41.6,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""high"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""none""}",True,tech_report,verified,Figure 44 · HealthBench Professional length-adjusted; metric=headline_metric. Exact printed value in the general-capability summary.
muse-spark-1.1,charxiv_reasoning,88.4,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""xhigh"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""Python""}",True,tech_report,verified,Figure 44 · CharXiv Reasoning w/code execution; metric=headline_metric. Exact printed value in the general-capability summary.
muse-spark-1.1,babyvision,76.3,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""xhigh"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""Python""}",True,tech_report,verified,Figure 44 · BabyVision w/code execution; metric=headline_metric. Exact printed value in the general-capability summary.
muse-spark,babyvision,39.9,https://ai.meta.com/static-resource/muse-spark-1.1-evaluation-report,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""default"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""Python""}",True,tech_report,verified,Figure 44 · BabyVision w/code execution; metric=headline_metric. Exact printed value in the general-capability summary.
claude-opus-4.8,meta_internal_coding_bench,69.0,https://ai.meta.com/blog/introducing-muse-spark-meta-model-api/,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""max"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""agentic coding""}",True,official_blog,verified,launch asset 04 · Meta Internal Coding Bench; metric=headline_metric. Exact numeric label visually read from the official launch raster.
muse-spark-1.1,meta_internal_coding_bench,68.3,https://ai.meta.com/blog/introducing-muse-spark-meta-model-api/,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""xhigh"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""agentic coding""}",True,official_blog,verified,launch asset 04 · Meta Internal Coding Bench; metric=headline_metric. Exact numeric label visually read from the official launch raster.
gpt-5.5,meta_internal_coding_bench,67.1,https://ai.meta.com/blog/introducing-muse-spark-meta-model-api/,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""xhigh"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""agentic coding""}",True,official_blog,verified,launch asset 04 · Meta Internal Coding Bench; metric=headline_metric. Exact numeric label visually read from the official launch raster.
gemini-3.1-pro,meta_internal_coding_bench,59.2,https://ai.meta.com/blog/introducing-muse-spark-meta-model-api/,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""high"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""agentic coding""}",True,official_blog,verified,launch asset 04 · Meta Internal Coding Bench; metric=headline_metric. Exact numeric label visually read from the official launch raster.
muse-spark,meta_internal_coding_bench,58.8,https://ai.meta.com/blog/introducing-muse-spark-meta-model-api/,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""default"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""agentic coding""}",True,official_blog,verified,launch asset 04 · Meta Internal Coding Bench; metric=headline_metric. Exact numeric label visually read from the official launch raster.
muse-spark-1.1,vibecodebench_v1_1_test,72.2,https://ai.meta.com/blog/introducing-muse-spark-meta-model-api/,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""xhigh"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""agentic coding""}",True,official_blog,verified,launch asset 09 · Vibe Code Bench v1.1 test; metric=headline_metric. Exact numeric label visually read from the official launch raster.
muse-spark,vibecodebench_v1_1_test,19.7,https://ai.meta.com/blog/introducing-muse-spark-meta-model-api/,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""default"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""agentic coding""}",True,official_blog,verified,launch asset 09 · Vibe Code Bench v1.1 test; metric=headline_metric. Exact numeric label visually read from the official launch raster.
muse-spark-1.1,swe_atlas_codebase_qna,42.0,https://ai.meta.com/blog/introducing-muse-spark-meta-model-api/,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""xhigh"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""codebase tools""}",True,official_blog,verified,launch asset 09 · SWE Atlas Codebase QnA; metric=headline_metric. Exact numeric label visually read from the official launch raster.
muse-spark,swe_atlas_codebase_qna,24.2,https://ai.meta.com/blog/introducing-muse-spark-meta-model-api/,"{""configuration"":""default"",""context"":""benchmark-specific"",""effort"":""default"",""harness"":""Meta MSL general-capability evaluation"",""mode"":""thinking"",""sampling"":""pass@1 unless source row states otherwise"",""temperature"":""1.0 for Muse; provider/source setting for comparisons"",""tools"":""codebase tools""}",True,official_blog,verified,launch asset 09 · SWE Atlas Codebase QnA; metric=headline_metric. Exact numeric label visually read from the official launch raster.
gemini-3-pro-deep-think-2025-12,hle,41.0,https://blog.google/products-and-platforms/products/gemini/gemini-3/#gemini-3-deep-think,"{""effort"":""vendor-specialized"",""harness"":""official Google / ARC Prize"",""judge"":""benchmark-specified"",""mode"":""deep-think"",""sampling"":""pass@1"",""snapshot"":""Nov-Dec 2025"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,Google Gemini 3 launch: Humanity's Last Exam: no tools. Distinct Dec 2025 Deep Think evaluated configuration.
gemini-3-pro-deep-think-2025-12,gpqa_diamond,93.8,https://blog.google/products-and-platforms/products/gemini/gemini-3/#gemini-3-deep-think,"{""effort"":""vendor-specialized"",""harness"":""official Google / ARC Prize"",""judge"":""benchmark-specified"",""mode"":""deep-think"",""sampling"":""pass@1"",""snapshot"":""Nov-Dec 2025"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,Google Gemini 3 launch: GPQA Diamond. Distinct Dec 2025 Deep Think evaluated configuration.
gemini-3-pro-deep-think-2025-12,arc_agi_2,45.1,https://blog.google/products-and-platforms/products/gemini/gemini-3/#gemini-3-deep-think,"{""effort"":""vendor-specialized"",""harness"":""official Google / ARC Prize"",""judge"":""benchmark-specified"",""mode"":""deep-think"",""sampling"":""pass@1"",""snapshot"":""Nov-Dec 2025"",""temperature"":""default"",""tools"":""code execution""}",False,official_blog,verified,Google Gemini 3 launch: ARC-AGI-2: code execution. Distinct Dec 2025 Deep Think evaluated configuration.
gemini-3.1-pro-deep-think-2026-02,arc_agi_2,84.6,https://deepmind.google/models/gemini/deep-think/,"{""effort"":""vendor-specialized"",""harness"":""Google Deep Think evaluation"",""judge"":""benchmark-specified"",""mode"":""deep-think"",""sampling"":""pass@1"",""snapshot"":""Feb 2026"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,Google February Deep Think table: ARC-AGI-2. ARC Prize Verified; v2 semi-private. Non-default sampling or Google cross-model harness is retained in reported_setting rather than changing benchmark identity.
gemini-3.1-pro-deep-think-2026-02,hle,48.4,https://deepmind.google/models/gemini/deep-think/,"{""effort"":""vendor-specialized"",""harness"":""Google Deep Think evaluation"",""judge"":""benchmark-specified"",""mode"":""deep-think"",""sampling"":""pass@1"",""snapshot"":""Feb 2026"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,"Google February Deep Think table: Humanity's Last Exam: full text+MM, no tools. Full 2,500-question text+multimodal set; no tools. Non-default sampling or Google cross-model harness is retained in reported_setting rather than changing benchmark identity."
gemini-3.1-pro-deep-think-2026-02,hle_tools,53.4,https://deepmind.google/models/gemini/deep-think/,"{""effort"":""vendor-specialized"",""harness"":""Google Deep Think evaluation"",""judge"":""benchmark-specified"",""mode"":""deep-think"",""sampling"":""pass@1"",""snapshot"":""Feb 2026"",""temperature"":""default"",""tools"":""search + code execution""}",False,official_blog,verified,Google February Deep Think table: Humanity's Last Exam: search + code execution. Full text+multimodal set; search and code execution. Non-default sampling or Google cross-model harness is retained in reported_setting rather than changing benchmark identity.
gemini-3.1-pro-deep-think-2026-02,mmmu_pro,81.5,https://deepmind.google/models/gemini/deep-think/,"{""effort"":""vendor-specialized"",""harness"":""Google Deep Think evaluation"",""judge"":""benchmark-specified"",""mode"":""deep-think"",""sampling"":""pass@1"",""snapshot"":""Feb 2026"",""temperature"":""default"",""tools"":""none""}",True,official_blog,verified,Google February Deep Think table: MMMU-Pro: Standard 10-option + Vision average. No tools. Non-default sampling or Google cross-model harness is retained in reported_setting rather than changing benchmark identity.
gemini-3.1-pro-deep-think-2026-02,imo_2025,81.5,https://deepmind.google/models/gemini/deep-think/,"{""effort"":""vendor-specialized"",""harness"":""Google Deep Think evaluation"",""judge"":""benchmark-specified"",""mode"":""deep-think"",""sampling"":""trials=4 (average)"",""snapshot"":""Feb 2026"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Google February Deep Think table: International Math Olympiad 2025. All 6 problems; average across 4 runs; expert grading. Non-default sampling or Google cross-model harness is retained in reported_setting rather than changing benchmark identity.
gemini-3-pro,imo_2025,14.3,https://deepmind.google/models/gemini/deep-think/,"{""effort"":""high"",""harness"":""Google self-computed cross-model evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""sampling"":""trials=4 (average)"",""snapshot"":""Feb 2026"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Google February Deep Think table: International Math Olympiad 2025. All 6 problems; average across 4 runs; expert grading. Non-default sampling or Google cross-model harness is retained in reported_setting rather than changing benchmark identity.
gpt-5.2,imo_2025,71.4,https://deepmind.google/models/gemini/deep-think/,"{""effort"":""xhigh"",""harness"":""Google self-computed cross-model evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""sampling"":""trials=4 (average)"",""snapshot"":""Feb 2026"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Google February Deep Think table: International Math Olympiad 2025. All 6 problems; average across 4 runs; expert grading. Non-default sampling or Google cross-model harness is retained in reported_setting rather than changing benchmark identity.
gemini-3.1-pro-deep-think-2026-02,ipho_2025_theory,87.7,https://deepmind.google/models/gemini/deep-think/,"{""effort"":""vendor-specialized"",""harness"":""Google Deep Think evaluation"",""judge"":""benchmark-specified"",""mode"":""deep-think"",""sampling"":""trials=8 (average)"",""snapshot"":""Feb 2026"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Google February Deep Think table: International Physics Olympiad 2025 Theory. All 3 used problems; average across 8 runs; Gemini judge. Non-default sampling or Google cross-model harness is retained in reported_setting rather than changing benchmark identity.
gemini-3-pro,ipho_2025_theory,76.3,https://deepmind.google/models/gemini/deep-think/,"{""effort"":""high"",""harness"":""Google self-computed cross-model evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""sampling"":""trials=8 (average)"",""snapshot"":""Feb 2026"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Google February Deep Think table: International Physics Olympiad 2025 Theory. All 3 used problems; average across 8 runs; Gemini judge. Non-default sampling or Google cross-model harness is retained in reported_setting rather than changing benchmark identity.
claude-opus-4.6,ipho_2025_theory,71.6,https://deepmind.google/models/gemini/deep-think/,"{""effort"":""max"",""harness"":""Google self-computed cross-model evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""sampling"":""trials=8 (average)"",""snapshot"":""Feb 2026"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Google February Deep Think table: International Physics Olympiad 2025 Theory. All 3 used problems; average across 8 runs; Gemini judge. Non-default sampling or Google cross-model harness is retained in reported_setting rather than changing benchmark identity.
gpt-5.2,ipho_2025_theory,70.5,https://deepmind.google/models/gemini/deep-think/,"{""effort"":""xhigh"",""harness"":""Google self-computed cross-model evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""sampling"":""trials=8 (average)"",""snapshot"":""Feb 2026"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Google February Deep Think table: International Physics Olympiad 2025 Theory. All 3 used problems; average across 8 runs; Gemini judge. Non-default sampling or Google cross-model harness is retained in reported_setting rather than changing benchmark identity.
gemini-3.1-pro-deep-think-2026-02,cmt_benchmark,50.5,https://deepmind.google/models/gemini/deep-think/,"{""effort"":""vendor-specialized"",""harness"":""Google Deep Think evaluation"",""judge"":""benchmark-specified"",""mode"":""deep-think"",""sampling"":""pass@8"",""snapshot"":""Feb 2026"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Google February Deep Think table: CMT-Benchmark. 50 problems; pass@8; Gemini judge. Non-default sampling or Google cross-model harness is retained in reported_setting rather than changing benchmark identity.
gemini-3-pro,cmt_benchmark,39.5,https://deepmind.google/models/gemini/deep-think/,"{""effort"":""high"",""harness"":""Google self-computed cross-model evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""sampling"":""pass@8"",""snapshot"":""Feb 2026"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Google February Deep Think table: CMT-Benchmark. 50 problems; pass@8; Gemini judge. Non-default sampling or Google cross-model harness is retained in reported_setting rather than changing benchmark identity.
claude-opus-4.6,cmt_benchmark,17.1,https://deepmind.google/models/gemini/deep-think/,"{""effort"":""max"",""harness"":""Google self-computed cross-model evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""sampling"":""pass@8"",""snapshot"":""Feb 2026"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Google February Deep Think table: CMT-Benchmark. 50 problems; pass@8; Gemini judge. Non-default sampling or Google cross-model harness is retained in reported_setting rather than changing benchmark identity.
gpt-5.2,cmt_benchmark,41.0,https://deepmind.google/models/gemini/deep-think/,"{""effort"":""xhigh"",""harness"":""Google self-computed cross-model evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""sampling"":""pass@8"",""snapshot"":""Feb 2026"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Google February Deep Think table: CMT-Benchmark. 50 problems; pass@8; Gemini judge. Non-default sampling or Google cross-model harness is retained in reported_setting rather than changing benchmark identity.
gemini-3.1-pro-deep-think-2026-02,icho_2025_theory,82.8,https://deepmind.google/models/gemini/deep-think/,"{""effort"":""vendor-specialized"",""harness"":""Google Deep Think evaluation"",""judge"":""benchmark-specified"",""mode"":""deep-think"",""sampling"":""trials=8 (average)"",""snapshot"":""Feb 2026"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Google February Deep Think table: International Chemistry Olympiad 2025 Theory. All 9 problems; average across 8 runs; Gemini judge. Non-default sampling or Google cross-model harness is retained in reported_setting rather than changing benchmark identity.
gemini-3-pro,icho_2025_theory,69.6,https://deepmind.google/models/gemini/deep-think/,"{""effort"":""high"",""harness"":""Google self-computed cross-model evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""sampling"":""trials=8 (average)"",""snapshot"":""Feb 2026"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Google February Deep Think table: International Chemistry Olympiad 2025 Theory. All 9 problems; average across 8 runs; Gemini judge. Non-default sampling or Google cross-model harness is retained in reported_setting rather than changing benchmark identity.
gpt-5.2,icho_2025_theory,72.0,https://deepmind.google/models/gemini/deep-think/,"{""effort"":""xhigh"",""harness"":""Google self-computed cross-model evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""sampling"":""trials=8 (average)"",""snapshot"":""Feb 2026"",""temperature"":""default"",""tools"":""none""}",False,official_blog,verified,Google February Deep Think table: International Chemistry Olympiad 2025 Theory. All 9 problems; average across 8 runs; Gemini judge. Non-default sampling or Google cross-model harness is retained in reported_setting rather than changing benchmark identity.
gemini-3-flash,terminal_bench_2_1,58.0,https://deepmind.google/models/model-cards/gemini-3-5-flash/,"{""effort"":""source does not state"",""harness"":""Terminus-2; Gemini self-computed, others public leaderboard"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""May 2026"",""temperature"":""default"",""tools"":""terminal agent""}",False,official_model_card,verified,Google Gemini 3.5 Flash May 2026 table: Terminal-Bench 2.1; Terminus-2 harness. Settings and provenance are preserved per cell.
gemini-3.5-flash,mcpatlas,83.6,https://deepmind.google/models/model-cards/gemini-3-5-flash/,"{""effort"":""medium (runtime default)"",""harness"":""ScaleAI official May 2026 leaderboard"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""May 2026"",""temperature"":""default"",""tools"":""MCP server tool use""}",True,official_model_card,verified,Google Gemini 3.5 Flash May 2026 table: MCP Atlas; ScaleAI official leaderboard. Settings and provenance are preserved per cell.
gemini-3.5-flash,toolathlon,56.5,https://deepmind.google/models/model-cards/gemini-3-5-flash/,"{""effort"":""medium (runtime default)"",""harness"":""benchmark-author supplied for Gemini"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""May 2026"",""temperature"":""default"",""tools"":""Toolathlon agent harness""}",True,official_model_card,verified,Google Gemini 3.5 Flash May 2026 table: Toolathlon; original 108-task benchmark. Settings and provenance are preserved per cell.
gemini-3-flash,osworld_verified,65.1,https://deepmind.google/models/model-cards/gemini-3-5-flash/,"{""effort"":""source does not state"",""harness"":""OSWorld Docker, 1080p, max 100 steps"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""Gemini avg 5 runs, single attempt per run"",""snapshot"":""May 2026"",""temperature"":""default"",""tools"":""pyautogui + UI-specific function declarations""}",False,official_model_card,verified,"Google Gemini 3.5 Flash May 2026 table: OSWorld-Verified; Docker, 1080p, max 100 steps. Settings and provenance are preserved per cell."
claude-opus-4.7,osworld_verified,78.0,https://deepmind.google/models/model-cards/gemini-3-5-flash/,"{""effort"":""max/best available"",""harness"":""OSWorld Docker, 1080p, max 100 steps"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""Gemini avg 5 runs, single attempt per run"",""snapshot"":""May 2026"",""temperature"":""default"",""tools"":""pyautogui + UI-specific function declarations""}",False,official_model_card,verified,"Google Gemini 3.5 Flash May 2026 table: OSWorld-Verified; Docker, 1080p, max 100 steps. Settings and provenance are preserved per cell."
gemini-3.5-flash,finance_agent_v2,57.9,https://deepmind.google/models/model-cards/gemini-3-5-flash/,"{""effort"":""medium (runtime default)"",""harness"":""Vals Finance Agent v2 leaderboard"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""mean of 3 runs"",""snapshot"":""May 2026"",""temperature"":""default"",""tools"":""Vals six-tool harness""}",True,official_model_card,verified,"Google Gemini 3.5 Flash May 2026 table: Finance Agent v2; Vals held-out leaderboard, weighted partial credit. Settings and provenance are preserved per cell."
gemini-3-flash,finance_agent_v2,42.6,https://deepmind.google/models/model-cards/gemini-3-5-flash/,"{""effort"":""source does not state"",""harness"":""Vals Finance Agent v2 leaderboard"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""mean of 3 runs"",""snapshot"":""May 2026"",""temperature"":""default"",""tools"":""Vals six-tool harness""}",True,official_model_card,verified,"Google Gemini 3.5 Flash May 2026 table: Finance Agent v2; Vals held-out leaderboard, weighted partial credit. Settings and provenance are preserved per cell."
claude-sonnet-4.6,finance_agent_v2,51.0,https://deepmind.google/models/model-cards/gemini-3-5-flash/,"{""effort"":""max/best available"",""harness"":""Vals Finance Agent v2 leaderboard"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""mean of 3 runs"",""snapshot"":""May 2026"",""temperature"":""default"",""tools"":""Vals six-tool harness""}",True,official_model_card,verified,"Google Gemini 3.5 Flash May 2026 table: Finance Agent v2; Vals held-out leaderboard, weighted partial credit. Settings and provenance are preserved per cell."
claude-opus-4.7,finance_agent_v2,51.5,https://deepmind.google/models/model-cards/gemini-3-5-flash/,"{""effort"":""max/best available"",""harness"":""Vals Finance Agent v2 leaderboard"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""mean of 3 runs"",""snapshot"":""May 2026"",""temperature"":""default"",""tools"":""Vals six-tool harness""}",True,official_model_card,verified,"Google Gemini 3.5 Flash May 2026 table: Finance Agent v2; Vals held-out leaderboard, weighted partial credit. Settings and provenance are preserved per cell."
gemini-3.5-flash,gdpval_aa_v2_elo,1656.0,https://deepmind.google/models/model-cards/gemini-3-5-flash/,"{""effort"":""medium (runtime default)"",""harness"":""Artificial Analysis GDPval-AA v2 leaderboard"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""May 2026"",""temperature"":""default"",""tools"":""Stirrup/E2B agent sandbox""}",True,official_model_card,verified,Google Gemini 3.5 Flash May 2026 table: GDPval-AA v2; Artificial Analysis Elo. Settings and provenance are preserved per cell.
gemini-3-flash,gdpval_aa_v2_elo,1204.0,https://deepmind.google/models/model-cards/gemini-3-5-flash/,"{""effort"":""source does not state"",""harness"":""Artificial Analysis GDPval-AA v2 leaderboard"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""May 2026"",""temperature"":""default"",""tools"":""Stirrup/E2B agent sandbox""}",True,official_model_card,verified,Google Gemini 3.5 Flash May 2026 table: GDPval-AA v2; Artificial Analysis Elo. Settings and provenance are preserved per cell.
claude-sonnet-4.6,gdpval_aa_v2_elo,1676.0,https://deepmind.google/models/model-cards/gemini-3-5-flash/,"{""effort"":""max/best available"",""harness"":""Artificial Analysis GDPval-AA v2 leaderboard"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""May 2026"",""temperature"":""default"",""tools"":""Stirrup/E2B agent sandbox""}",True,official_model_card,verified,Google Gemini 3.5 Flash May 2026 table: GDPval-AA v2; Artificial Analysis Elo. Settings and provenance are preserved per cell.
claude-opus-4.7,gdpval_aa_v2_elo,1753.0,https://deepmind.google/models/model-cards/gemini-3-5-flash/,"{""effort"":""max/best available"",""harness"":""Artificial Analysis GDPval-AA v2 leaderboard"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""May 2026"",""temperature"":""default"",""tools"":""Stirrup/E2B agent sandbox""}",True,official_model_card,verified,Google Gemini 3.5 Flash May 2026 table: GDPval-AA v2; Artificial Analysis Elo. Settings and provenance are preserved per cell.
gemini-3.5-flash,charxiv_reasoning,84.2,https://deepmind.google/models/model-cards/gemini-3-5-flash/,"{""effort"":""medium (runtime default)"",""harness"":""Google self-computed for Gemini/GPT; provider for Claude"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""May 2026"",""temperature"":""default"",""tools"":""none""}",False,official_model_card,verified,Google Gemini 3.5 Flash May 2026 table: CharXiv Reasoning; no tools. Settings and provenance are preserved per cell.
gemini-3.5-flash,mmmu_pro,83.6,https://deepmind.google/models/model-cards/gemini-3-5-flash/,"{""effort"":""medium (runtime default)"",""harness"":""Standard 10-option + Vision arithmetic mean"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""May 2026"",""temperature"":""default"",""tools"":""none""}",True,official_model_card,verified,Google Gemini 3.5 Flash May 2026 table: MMMU-Pro; Standard 10-option + Vision average; no tools. Settings and provenance are preserved per cell.
claude-opus-4.7,mmmu_pro,75.2,https://deepmind.google/models/model-cards/gemini-3-5-flash/,"{""effort"":""max/best available"",""harness"":""Standard 10-option + Vision arithmetic mean"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""May 2026"",""temperature"":""default"",""tools"":""none""}",True,official_model_card,verified,Google Gemini 3.5 Flash May 2026 table: MMMU-Pro; Standard 10-option + Vision average; no tools. Settings and provenance are preserved per cell.
gemini-3.5-flash,blueprint_bench_2,33.6,https://deepmind.google/models/model-cards/gemini-3-5-flash/,"{""effort"":""medium (runtime default)"",""harness"":""Andon Labs public leaderboard"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""May 2026"",""temperature"":""default"",""tools"":""persistent cross-apartment notepad""}",False,official_model_card,verified,Google Gemini 3.5 Flash May 2026 table: Blueprint-Bench 2; normalized 0-100 leaderboard score. Settings and provenance are preserved per cell.
gemini-3-flash,blueprint_bench_2,0.0,https://deepmind.google/models/model-cards/gemini-3-5-flash/,"{""effort"":""source does not state"",""harness"":""Andon Labs public leaderboard"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""May 2026"",""temperature"":""default"",""tools"":""persistent cross-apartment notepad""}",False,official_model_card,verified,Google Gemini 3.5 Flash May 2026 table: Blueprint-Bench 2; normalized 0-100 leaderboard score. Settings and provenance are preserved per cell.
gemini-3.1-pro,blueprint_bench_2,26.5,https://deepmind.google/models/model-cards/gemini-3-5-flash/,"{""effort"":""source does not state"",""harness"":""Andon Labs public leaderboard"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""May 2026"",""temperature"":""default"",""tools"":""persistent cross-apartment notepad""}",False,official_model_card,verified,Google Gemini 3.5 Flash May 2026 table: Blueprint-Bench 2; normalized 0-100 leaderboard score. Settings and provenance are preserved per cell.
claude-sonnet-4.6,blueprint_bench_2,6.7,https://deepmind.google/models/model-cards/gemini-3-5-flash/,"{""effort"":""max/best available"",""harness"":""Andon Labs public leaderboard"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""May 2026"",""temperature"":""default"",""tools"":""persistent cross-apartment notepad""}",False,official_model_card,verified,Google Gemini 3.5 Flash May 2026 table: Blueprint-Bench 2; normalized 0-100 leaderboard score. Settings and provenance are preserved per cell.
claude-opus-4.7,blueprint_bench_2,24.5,https://deepmind.google/models/model-cards/gemini-3-5-flash/,"{""effort"":""max/best available"",""harness"":""Andon Labs public leaderboard"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""May 2026"",""temperature"":""default"",""tools"":""persistent cross-apartment notepad""}",False,official_model_card,verified,Google Gemini 3.5 Flash May 2026 table: Blueprint-Bench 2; normalized 0-100 leaderboard score. Settings and provenance are preserved per cell.
gpt-5.5,blueprint_bench_2,36.2,https://deepmind.google/models/model-cards/gemini-3-5-flash/,"{""effort"":""xhigh/best available"",""harness"":""Andon Labs public leaderboard"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""May 2026"",""temperature"":""default"",""tools"":""persistent cross-apartment notepad""}",False,official_model_card,verified,Google Gemini 3.5 Flash May 2026 table: Blueprint-Bench 2; normalized 0-100 leaderboard score. Settings and provenance are preserved per cell.
gemini-3.5-flash,gdm_mrcr_v2_8needle_upto_128k,77.3,https://deepmind.google/models/model-cards/gemini-3-5-flash/,"{""effort"":""medium (runtime default)"",""harness"":""Google DeepMind eval_hub MRCR v2 scorer"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""May 2026"",""temperature"":""default"",""tools"":""none""}",True,official_model_card,verified,Google Gemini 3.5 Flash May 2026 table: GDM MRCR v2 8-needle: 128k average; cumulative upto_128K. Settings and provenance are preserved per cell.
gemini-3-flash,gdm_mrcr_v2_8needle_upto_128k,67.2,https://deepmind.google/models/model-cards/gemini-3-5-flash/,"{""effort"":""source does not state"",""harness"":""Google DeepMind eval_hub MRCR v2 scorer"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""May 2026"",""temperature"":""default"",""tools"":""none""}",True,official_model_card,verified,Google Gemini 3.5 Flash May 2026 table: GDM MRCR v2 8-needle: 128k average; cumulative upto_128K. Settings and provenance are preserved per cell.
gemini-3.1-pro,gdm_mrcr_v2_8needle_upto_128k,84.9,https://deepmind.google/models/model-cards/gemini-3-5-flash/,"{""effort"":""source does not state"",""harness"":""Google DeepMind eval_hub MRCR v2 scorer"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""May 2026"",""temperature"":""default"",""tools"":""none""}",True,official_model_card,verified,Google Gemini 3.5 Flash May 2026 table: GDM MRCR v2 8-needle: 128k average; cumulative upto_128K. Settings and provenance are preserved per cell.
claude-sonnet-4.6,gdm_mrcr_v2_8needle_upto_128k,84.9,https://deepmind.google/models/model-cards/gemini-3-5-flash/,"{""effort"":""max/best available"",""harness"":""Google DeepMind eval_hub MRCR v2 scorer"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""May 2026"",""temperature"":""default"",""tools"":""none""}",True,official_model_card,verified,Google Gemini 3.5 Flash May 2026 table: GDM MRCR v2 8-needle: 128k average; cumulative upto_128K. Settings and provenance are preserved per cell.
claude-opus-4.7,gdm_mrcr_v2_8needle_upto_128k,59.3,https://deepmind.google/models/model-cards/gemini-3-5-flash/,"{""effort"":""max/best available"",""harness"":""Google DeepMind eval_hub MRCR v2 scorer"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""May 2026"",""temperature"":""default"",""tools"":""none""}",True,official_model_card,verified,Google Gemini 3.5 Flash May 2026 table: GDM MRCR v2 8-needle: 128k average; cumulative upto_128K. Settings and provenance are preserved per cell.
gpt-5.5,gdm_mrcr_v2_8needle_upto_128k,94.8,https://deepmind.google/models/model-cards/gemini-3-5-flash/,"{""effort"":""xhigh/best available"",""harness"":""Google DeepMind eval_hub MRCR v2 scorer"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""May 2026"",""temperature"":""default"",""tools"":""none""}",True,official_model_card,verified,Google Gemini 3.5 Flash May 2026 table: GDM MRCR v2 8-needle: 128k average; cumulative upto_128K. Settings and provenance are preserved per cell.
gemini-3.5-flash,gdm_mrcr_v2_8needle_1m,26.6,https://deepmind.google/models/model-cards/gemini-3-5-flash/,"{""effort"":""medium (runtime default)"",""harness"":""Google DeepMind eval_hub MRCR v2 scorer"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""May 2026"",""temperature"":""default"",""tools"":""none""}",True,official_model_card,verified,"Google Gemini 3.5 Flash May 2026 table: GDM MRCR v2 8-needle: 1M pointwise; single (524288,1048576] context bin. Settings and provenance are preserved per cell."
gemini-3-flash,gdm_mrcr_v2_8needle_1m,22.1,https://deepmind.google/models/model-cards/gemini-3-5-flash/,"{""effort"":""source does not state"",""harness"":""Google DeepMind eval_hub MRCR v2 scorer"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""May 2026"",""temperature"":""default"",""tools"":""none""}",True,official_model_card,verified,"Google Gemini 3.5 Flash May 2026 table: GDM MRCR v2 8-needle: 1M pointwise; single (524288,1048576] context bin. Settings and provenance are preserved per cell."
gemini-3.1-pro,gdm_mrcr_v2_8needle_1m,26.3,https://deepmind.google/models/model-cards/gemini-3-5-flash/,"{""effort"":""source does not state"",""harness"":""Google DeepMind eval_hub MRCR v2 scorer"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""May 2026"",""temperature"":""default"",""tools"":""none""}",True,official_model_card,verified,"Google Gemini 3.5 Flash May 2026 table: GDM MRCR v2 8-needle: 1M pointwise; single (524288,1048576] context bin. Settings and provenance are preserved per cell."
gemini-3.5-flash,arc_agi_2,72.1,https://deepmind.google/models/model-cards/gemini-3-5-flash/,"{""effort"":""medium (runtime default)"",""harness"":""ARC Prize verified leaderboard"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""May 2026"",""temperature"":""default"",""tools"":""none""}",True,official_model_card,verified,Google Gemini 3.5 Flash May 2026 table: ARC-AGI-2; ARC Prize Verified; semi-private. Settings and provenance are preserved per cell.
gemini-3.5-flash-lite,swe_bench_pro,54.2,https://deepmind.google/models/model-cards/gemini-3-5-flash-lite/,"{""effort"":""high"",""harness"":""Google internal Antigravity for Gemini; provider for others"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""code execution""}",False,official_model_card,verified,"Google Gemini 3.5 Flash-Lite July 2026 matrix: SWE-Bench Pro (Public). Exact source effort, tools, sampling and harness are preserved."
gemini-3.1-flash-lite,swe_bench_pro,38.3,https://deepmind.google/models/model-cards/gemini-3-5-flash-lite/,"{""effort"":""high"",""harness"":""Google internal Antigravity for Gemini; provider for others"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""code execution""}",False,official_model_card,verified,"Google Gemini 3.5 Flash-Lite July 2026 matrix: SWE-Bench Pro (Public). Exact source effort, tools, sampling and harness are preserved."
gpt-5.4-mini,swe_bench_pro,54.4,https://deepmind.google/models/model-cards/gemini-3-5-flash-lite/,"{""effort"":""xhigh/best available"",""harness"":""Google internal Antigravity for Gemini; provider for others"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""code execution""}",False,official_model_card,verified,"Google Gemini 3.5 Flash-Lite July 2026 matrix: SWE-Bench Pro (Public). Exact source effort, tools, sampling and harness are preserved."
gemini-3.5-flash-lite,terminal_bench_2_1,54.0,https://deepmind.google/models/model-cards/gemini-3-5-flash-lite/,"{""effort"":""high"",""harness"":""Terminus-2"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""terminal agent""}",False,official_model_card,verified,"Google Gemini 3.5 Flash-Lite July 2026 matrix: Terminal-Bench 2.1. Exact source effort, tools, sampling and harness are preserved."
gemini-3.1-flash-lite,terminal_bench_2_1,31.0,https://deepmind.google/models/model-cards/gemini-3-5-flash-lite/,"{""effort"":""high"",""harness"":""Terminus-2"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""terminal agent""}",False,official_model_card,verified,"Google Gemini 3.5 Flash-Lite July 2026 matrix: Terminal-Bench 2.1. Exact source effort, tools, sampling and harness are preserved."
gpt-5.4-mini,terminal_bench_2_1,59.2,https://deepmind.google/models/model-cards/gemini-3-5-flash-lite/,"{""effort"":""xhigh/best available"",""harness"":""Terminus-2"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""terminal agent""}",False,official_model_card,verified,"Google Gemini 3.5 Flash-Lite July 2026 matrix: Terminal-Bench 2.1. Exact source effort, tools, sampling and harness are preserved."
claude-haiku-4.5,terminal_bench_2_1,44.2,https://deepmind.google/models/model-cards/gemini-3-5-flash-lite/,"{""effort"":""extended thinking/best available"",""harness"":""Terminus-2"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""terminal agent""}",False,official_model_card,verified,"Google Gemini 3.5 Flash-Lite July 2026 matrix: Terminal-Bench 2.1. Exact source effort, tools, sampling and harness are preserved."
gemini-3.5-flash-lite,mle_bench_partial30_avg_position_k2,39.2,https://deepmind.google/models/model-cards/gemini-3-5-flash-lite/,"{""effort"":""high"",""harness"":""Google MLE-Bench agent harness"",""judge"":""Kaggle private-leaderboard rank"",""mode"":""thinking/reasoning"",""sampling"":""2 independent runs/problem"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""interactive Bash + internet + H100""}",False,official_model_card,verified,"Google Gemini 3.5 Flash-Lite July 2026 matrix: MLE-Bench Partial 30: Average Position Score, k=2. Exact source effort, tools, sampling and harness are preserved."
gemini-3.1-flash-lite,mle_bench_partial30_avg_position_k2,22.0,https://deepmind.google/models/model-cards/gemini-3-5-flash-lite/,"{""effort"":""high"",""harness"":""Google MLE-Bench agent harness"",""judge"":""Kaggle private-leaderboard rank"",""mode"":""thinking/reasoning"",""sampling"":""2 independent runs/problem"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""interactive Bash + internet + H100""}",False,official_model_card,verified,"Google Gemini 3.5 Flash-Lite July 2026 matrix: MLE-Bench Partial 30: Average Position Score, k=2. Exact source effort, tools, sampling and harness are preserved."
gemini-3.5-flash-lite,gdpval_aa_v2_elo,1140.0,https://deepmind.google/models/model-cards/gemini-3-5-flash-lite/,"{""effort"":""high"",""harness"":""Artificial Analysis GDPval-AA v2 leaderboard"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""Stirrup/E2B agent sandbox""}",False,official_model_card,verified,"Google Gemini 3.5 Flash-Lite July 2026 matrix: GDPval-AA v2 Elo. Exact source effort, tools, sampling and harness are preserved."
gemini-3.1-flash-lite,gdpval_aa_v2_elo,642.0,https://deepmind.google/models/model-cards/gemini-3-5-flash-lite/,"{""effort"":""high"",""harness"":""Artificial Analysis GDPval-AA v2 leaderboard"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""Stirrup/E2B agent sandbox""}",False,official_model_card,verified,"Google Gemini 3.5 Flash-Lite July 2026 matrix: GDPval-AA v2 Elo. Exact source effort, tools, sampling and harness are preserved."
gpt-5.4-mini,gdpval_aa_v2_elo,1171.0,https://deepmind.google/models/model-cards/gemini-3-5-flash-lite/,"{""effort"":""xhigh/best available"",""harness"":""Artificial Analysis GDPval-AA v2 leaderboard"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""Stirrup/E2B agent sandbox""}",False,official_model_card,verified,"Google Gemini 3.5 Flash-Lite July 2026 matrix: GDPval-AA v2 Elo. Exact source effort, tools, sampling and harness are preserved."
claude-haiku-4.5,gdpval_aa_v2_elo,907.0,https://deepmind.google/models/model-cards/gemini-3-5-flash-lite/,"{""effort"":""extended thinking/best available"",""harness"":""Artificial Analysis GDPval-AA v2 leaderboard"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""Stirrup/E2B agent sandbox""}",True,official_model_card,verified,"Google Gemini 3.5 Flash-Lite July 2026 matrix: GDPval-AA v2 Elo. Exact source effort, tools, sampling and harness are preserved."
gemini-3.5-flash-lite,osworld_verified,74.0,https://deepmind.google/models/model-cards/gemini-3-5-flash-lite/,"{""effort"":""high"",""harness"":""official OSWorld Docker, 1080p, max 100 steps"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""Gemini avg 5 runs; single attempt/run"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""pyautogui + UI-specific functions""}",False,official_model_card,verified,"Google Gemini 3.5 Flash-Lite July 2026 matrix: OSWorld-Verified. Exact source effort, tools, sampling and harness are preserved."
gemini-3.1-flash-lite,osworld_verified,54.3,https://deepmind.google/models/model-cards/gemini-3-5-flash-lite/,"{""effort"":""high"",""harness"":""official OSWorld Docker, 1080p, max 100 steps"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""Gemini avg 5 runs; single attempt/run"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""pyautogui + UI-specific functions""}",False,official_model_card,verified,"Google Gemini 3.5 Flash-Lite July 2026 matrix: OSWorld-Verified. Exact source effort, tools, sampling and harness are preserved."
gpt-5.4-mini,osworld_verified,72.1,https://deepmind.google/models/model-cards/gemini-3-5-flash-lite/,"{""effort"":""xhigh/best available"",""harness"":""official OSWorld Docker, 1080p, max 100 steps"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""Gemini avg 5 runs; single attempt/run"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""pyautogui + UI-specific functions""}",False,official_model_card,verified,"Google Gemini 3.5 Flash-Lite July 2026 matrix: OSWorld-Verified. Exact source effort, tools, sampling and harness are preserved."
claude-haiku-4.5,osworld_verified,50.7,https://deepmind.google/models/model-cards/gemini-3-5-flash-lite/,"{""effort"":""extended thinking/best available"",""harness"":""official OSWorld Docker, 1080p, max 100 steps"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""Gemini avg 5 runs; single attempt/run"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""pyautogui + UI-specific functions""}",False,official_model_card,verified,"Google Gemini 3.5 Flash-Lite July 2026 matrix: OSWorld-Verified. Exact source effort, tools, sampling and harness are preserved."
gemini-3.5-flash-lite,charxiv_reasoning,74.5,https://deepmind.google/models/model-cards/gemini-3-5-flash-lite/,"{""effort"":""high"",""harness"":""Google self-computed"",""judge"":""official CharXiv binary answer judge"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""none""}",False,official_model_card,verified,"Google Gemini 3.5 Flash-Lite July 2026 matrix: CharXiv Reasoning: no tools. Exact source effort, tools, sampling and harness are preserved."
gemini-3.1-flash-lite,charxiv_reasoning,73.2,https://deepmind.google/models/model-cards/gemini-3-5-flash-lite/,"{""effort"":""high"",""harness"":""Google self-computed"",""judge"":""official CharXiv binary answer judge"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""none""}",False,official_model_card,verified,"Google Gemini 3.5 Flash-Lite July 2026 matrix: CharXiv Reasoning: no tools. Exact source effort, tools, sampling and harness are preserved."
gpt-5.4-mini,charxiv_reasoning,80.3,https://deepmind.google/models/model-cards/gemini-3-5-flash-lite/,"{""effort"":""xhigh/best available"",""harness"":""Google self-computed"",""judge"":""official CharXiv binary answer judge"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""none""}",False,official_model_card,verified,"Google Gemini 3.5 Flash-Lite July 2026 matrix: CharXiv Reasoning: no tools. Exact source effort, tools, sampling and harness are preserved."
claude-haiku-4.5,charxiv_reasoning,61.7,https://deepmind.google/models/model-cards/gemini-3-5-flash-lite/,"{""effort"":""extended thinking/best available"",""harness"":""Google self-computed"",""judge"":""official CharXiv binary answer judge"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""none""}",True,official_model_card,verified,"Google Gemini 3.5 Flash-Lite July 2026 matrix: CharXiv Reasoning: no tools. Exact source effort, tools, sampling and harness are preserved."
gemini-3.5-flash-lite,gdm_mrcr_v2_8needle_upto_128k,72.2,https://deepmind.google/models/model-cards/gemini-3-5-flash-lite/,"{""effort"":""high"",""harness"":""Google DeepMind eval_hub MRCR v2"",""judge"":""hash check + SequenceMatcher strict scorer"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""none""}",False,official_model_card,verified,"Google Gemini 3.5 Flash-Lite July 2026 matrix: GDM MRCR v2 8-needle: 128k average. Exact source effort, tools, sampling and harness are preserved."
gemini-3.1-flash-lite,gdm_mrcr_v2_8needle_upto_128k,60.1,https://deepmind.google/models/model-cards/gemini-3-5-flash-lite/,"{""effort"":""high"",""harness"":""Google DeepMind eval_hub MRCR v2"",""judge"":""hash check + SequenceMatcher strict scorer"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""none""}",False,official_model_card,verified,"Google Gemini 3.5 Flash-Lite July 2026 matrix: GDM MRCR v2 8-needle: 128k average. Exact source effort, tools, sampling and harness are preserved."
gpt-5.4-mini,gdm_mrcr_v2_8needle_upto_128k,42.7,https://deepmind.google/models/model-cards/gemini-3-5-flash-lite/,"{""effort"":""xhigh/best available"",""harness"":""Google DeepMind eval_hub MRCR v2"",""judge"":""hash check + SequenceMatcher strict scorer"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""none""}",False,official_model_card,verified,"Google Gemini 3.5 Flash-Lite July 2026 matrix: GDM MRCR v2 8-needle: 128k average. Exact source effort, tools, sampling and harness are preserved."
claude-haiku-4.5,gdm_mrcr_v2_8needle_upto_128k,35.3,https://deepmind.google/models/model-cards/gemini-3-5-flash-lite/,"{""effort"":""extended thinking/best available"",""harness"":""Google DeepMind eval_hub MRCR v2"",""judge"":""hash check + SequenceMatcher strict scorer"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""none""}",True,official_model_card,verified,"Google Gemini 3.5 Flash-Lite July 2026 matrix: GDM MRCR v2 8-needle: 128k average. Exact source effort, tools, sampling and harness are preserved."
gemini-3.5-flash-lite,gdm_mrcr_v2_8needle_1m,21.3,https://deepmind.google/models/model-cards/gemini-3-5-flash-lite/,"{""effort"":""high"",""harness"":""Google DeepMind eval_hub MRCR v2"",""judge"":""hash check + SequenceMatcher strict scorer"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""none""}",False,official_model_card,verified,"Google Gemini 3.5 Flash-Lite July 2026 matrix: GDM MRCR v2 8-needle: 1M pointwise. Exact source effort, tools, sampling and harness are preserved."
gemini-3.1-flash-lite,gdm_mrcr_v2_8needle_1m,12.3,https://deepmind.google/models/model-cards/gemini-3-5-flash-lite/,"{""effort"":""high"",""harness"":""Google DeepMind eval_hub MRCR v2"",""judge"":""hash check + SequenceMatcher strict scorer"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""none""}",False,official_model_card,verified,"Google Gemini 3.5 Flash-Lite July 2026 matrix: GDM MRCR v2 8-needle: 1M pointwise. Exact source effort, tools, sampling and harness are preserved."
gemini-3.6-flash,swe_bench_pro,58.7,https://deepmind.google/models/model-cards/gemini-3-6-flash/,"{""effort"":""medium (default)"",""harness"":""Google internal Antigravity for Gemini; provider for others"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""code execution""}",False,official_model_card,verified,"Google Gemini 3.6 Flash July 2026 matrix: SWE-Bench Pro (Public). Exact effort, tools, sampling and harness are preserved."
grok-4.5,swe_bench_pro,64.7,https://media.x.ai/v1/website/4p5-5184fdf9.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""fixed SWE-bench Pro agent scaffold"",""judge"":""verification reward and repository tests"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic repository shell/editor""}",True,model_card,verified,card45 SWE-Bench Pro Resolve rate (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
gemini-3.6-flash,deep_swe_v1_1,49.0,https://deepmind.google/models/model-cards/gemini-3-6-flash/,"{""effort"":""high"",""harness"":""DataCurve DeepSWE v1.1 leaderboard"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""code execution""}",False,official_model_card,verified,"Google Gemini 3.6 Flash July 2026 matrix: DeepSWE v1.1. Exact effort, tools, sampling and harness are preserved."
gemini-3.5-flash,deep_swe_v1_1,37.0,https://deepmind.google/models/model-cards/gemini-3-6-flash/,"{""effort"":""medium"",""harness"":""DataCurve DeepSWE v1.1 leaderboard"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""code execution""}",True,official_model_card,verified,"Google Gemini 3.6 Flash July 2026 matrix: DeepSWE v1.1. Exact effort, tools, sampling and harness are preserved."
grok-4.5,deep_swe_v1_1,54.0,https://x.ai/news/grok-4-6,"{""context"":""unknown"",""effort"":""high"",""harness"":""mini-swe-agent run by Datacurve"",""judge"":""isolated repository verifier"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic repository shell/editor""}",True,official_blog,verified,release46 DeepSWE v1.1 Pass@1 (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
gemini-3.6-flash,terminal_bench_2_1,78.0,https://deepmind.google/models/model-cards/gemini-3-6-flash/,"{""effort"":""medium (default)"",""harness"":""Terminus-2"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""terminal agent""}",False,official_model_card,verified,"Google Gemini 3.6 Flash July 2026 matrix: Terminal-Bench 2.1. Exact effort, tools, sampling and harness are preserved."
grok-4.5,terminal_bench_2_1,83.3,https://media.x.ai/v1/website/4p5-5184fdf9.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""Grok Build for Grok; source-reported peer harnesses"",""judge"":""verified task-success evaluator"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""container terminal""}",True,model_card,verified,card45 Terminal-Bench 2.1 Task success rate (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
gemini-3.6-flash,mle_bench_partial30_avg_position_k2,63.9,https://deepmind.google/models/model-cards/gemini-3-6-flash/,"{""effort"":""medium (default)"",""harness"":""Google MLE-Bench agent harness"",""judge"":""Kaggle private-leaderboard rank"",""mode"":""thinking/reasoning"",""sampling"":""2 independent runs/problem"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""interactive Bash + internet + H100""}",True,official_model_card,verified,"Google Gemini 3.6 Flash July 2026 matrix: MLE-Bench Partial 30: Average Position Score, k=2. Exact effort, tools, sampling and harness are preserved."
gemini-3.5-flash,mle_bench_partial30_avg_position_k2,49.7,https://deepmind.google/models/model-cards/gemini-3-6-flash/,"{""effort"":""medium (default)"",""harness"":""Google MLE-Bench agent harness"",""judge"":""Kaggle private-leaderboard rank"",""mode"":""thinking/reasoning"",""sampling"":""2 independent runs/problem"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""interactive Bash + internet + H100""}",True,official_model_card,verified,"Google Gemini 3.6 Flash July 2026 matrix: MLE-Bench Partial 30: Average Position Score, k=2. Exact effort, tools, sampling and harness are preserved."
gemini-3.1-pro,mle_bench_partial30_avg_position_k2,42.6,https://deepmind.google/models/model-cards/gemini-3-6-flash/,"{""effort"":""high/default"",""harness"":""Google MLE-Bench agent harness"",""judge"":""Kaggle private-leaderboard rank"",""mode"":""thinking/reasoning"",""sampling"":""2 independent runs/problem"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""interactive Bash + internet + H100""}",True,official_model_card,verified,"Google Gemini 3.6 Flash July 2026 matrix: MLE-Bench Partial 30: Average Position Score, k=2. Exact effort, tools, sampling and harness are preserved."
gpt-5.6-luna,mle_bench_partial30_avg_position_k2,47.6,https://deepmind.google/models/model-cards/gemini-3-6-flash/,"{""effort"":""maximum/best available"",""harness"":""Google MLE-Bench agent harness"",""judge"":""Kaggle private-leaderboard rank"",""mode"":""thinking/reasoning"",""sampling"":""2 independent runs/problem"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""interactive Bash + internet + H100""}",True,official_model_card,verified,"Google Gemini 3.6 Flash July 2026 matrix: MLE-Bench Partial 30: Average Position Score, k=2. Exact effort, tools, sampling and harness are preserved."
grok-4.5,mle_bench_partial30_avg_position_k2,43.2,https://deepmind.google/models/model-cards/gemini-3-6-flash/,"{""effort"":""maximum/best available"",""harness"":""Google MLE-Bench agent harness"",""judge"":""Kaggle private-leaderboard rank"",""mode"":""thinking/reasoning"",""sampling"":""2 independent runs/problem"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""interactive Bash + internet + H100""}",True,official_model_card,verified,"Google Gemini 3.6 Flash July 2026 matrix: MLE-Bench Partial 30: Average Position Score, k=2. Exact effort, tools, sampling and harness are preserved."
claude-sonnet-5,mle_bench_partial30_avg_position_k2,66.9,https://deepmind.google/models/model-cards/gemini-3-6-flash/,"{""effort"":""maximum/best available"",""harness"":""Google MLE-Bench agent harness"",""judge"":""Kaggle private-leaderboard rank"",""mode"":""thinking/reasoning"",""sampling"":""2 independent runs/problem"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""interactive Bash + internet + H100""}",True,official_model_card,verified,"Google Gemini 3.6 Flash July 2026 matrix: MLE-Bench Partial 30: Average Position Score, k=2. Exact effort, tools, sampling and harness are preserved."
gemini-3.6-flash,gdpval_aa_v2_elo,1421.0,https://deepmind.google/models/model-cards/gemini-3-6-flash/,"{""effort"":""medium (default)"",""harness"":""Artificial Analysis GDPval-AA v2 leaderboard"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""Stirrup/E2B agent sandbox""}",True,official_model_card,verified,"Google Gemini 3.6 Flash July 2026 matrix: GDPval-AA v2 Elo. Exact effort, tools, sampling and harness are preserved."
gpt-5.6-luna,gdpval_aa_v2_elo,1584.0,https://deepmind.google/models/model-cards/gemini-3-6-flash/,"{""effort"":""maximum/best available"",""harness"":""Artificial Analysis GDPval-AA v2 leaderboard"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""Stirrup/E2B agent sandbox""}",True,official_model_card,verified,"Google Gemini 3.6 Flash July 2026 matrix: GDPval-AA v2 Elo. Exact effort, tools, sampling and harness are preserved."
grok-4.5,gdpval_aa_v2_elo,1526.0,https://x.ai/news/grok-4-6,"{""context"":""unknown"",""effort"":""high"",""harness"":""Artificial Analysis GDPval-AA v2 harness"",""judge"":""pairwise Bradley-Terry Elo"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""benchmark-specified knowledge-work tools""}",True,official_blog,verified,release46 GDPVal-AA v2 Elo; source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
claude-sonnet-5,gdpval_aa_v2_elo,1607.0,https://deepmind.google/models/model-cards/gemini-3-6-flash/,"{""effort"":""maximum/best available"",""harness"":""Artificial Analysis GDPval-AA v2 leaderboard"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""Stirrup/E2B agent sandbox""}",True,official_model_card,verified,"Google Gemini 3.6 Flash July 2026 matrix: GDPval-AA v2 Elo. Exact effort, tools, sampling and harness are preserved."
gemini-3.6-flash,osworld_verified,83.0,https://deepmind.google/models/model-cards/gemini-3-6-flash/,"{""effort"":""medium (default)"",""harness"":""official OSWorld Docker, 1080p, max 100 steps"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""Google self-computed avg 5; single attempt/run"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""pyautogui + UI-specific functions""}",True,official_model_card,verified,"Google Gemini 3.6 Flash July 2026 matrix: OSWorld-Verified. Exact effort, tools, sampling and harness are preserved."
gpt-5.6-luna,osworld_verified,72.6,https://deepmind.google/models/model-cards/gemini-3-6-flash/,"{""effort"":""maximum/best available"",""harness"":""official OSWorld Docker, 1080p, max 100 steps"",""judge"":""benchmark-specified"",""mode"":""thinking/reasoning"",""sampling"":""Google self-computed avg 5; single attempt/run"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""pyautogui + UI-specific functions""}",True,official_model_card,verified,"Google Gemini 3.6 Flash July 2026 matrix: OSWorld-Verified. Exact effort, tools, sampling and harness are preserved."
gemini-3.6-flash,charxiv_reasoning,85.2,https://deepmind.google/models/model-cards/gemini-3-6-flash/,"{""effort"":""medium (default)"",""harness"":""Google self-computed for Gemini/Luna/Grok"",""judge"":""official CharXiv binary answer judge"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""none""}",True,official_model_card,verified,"Google Gemini 3.6 Flash July 2026 matrix: CharXiv Reasoning: no tools. Exact effort, tools, sampling and harness are preserved."
gpt-5.6-luna,charxiv_reasoning,82.7,https://deepmind.google/models/model-cards/gemini-3-6-flash/,"{""effort"":""maximum/best available"",""harness"":""Google self-computed for Gemini/Luna/Grok"",""judge"":""official CharXiv binary answer judge"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""none""}",True,official_model_card,verified,"Google Gemini 3.6 Flash July 2026 matrix: CharXiv Reasoning: no tools. Exact effort, tools, sampling and harness are preserved."
grok-4.5,charxiv_reasoning,81.6,https://deepmind.google/models/model-cards/gemini-3-6-flash/,"{""effort"":""maximum/best available"",""harness"":""Google self-computed for Gemini/Luna/Grok"",""judge"":""official CharXiv binary answer judge"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""none""}",True,official_model_card,verified,"Google Gemini 3.6 Flash July 2026 matrix: CharXiv Reasoning: no tools. Exact effort, tools, sampling and harness are preserved."
gemini-3.6-flash,gdm_mrcr_v2_8needle_upto_128k,91.8,https://deepmind.google/models/model-cards/gemini-3-6-flash/,"{""effort"":""medium (default)"",""harness"":""Google DeepMind eval_hub / Context Arena"",""judge"":""hash check + SequenceMatcher strict scorer"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""none""}",True,official_model_card,verified,"Google Gemini 3.6 Flash July 2026 matrix: GDM MRCR v2 8-needle: 128k average. Exact effort, tools, sampling and harness are preserved."
gpt-5.6-luna,gdm_mrcr_v2_8needle_upto_128k,74.8,https://deepmind.google/models/model-cards/gemini-3-6-flash/,"{""effort"":""maximum/best available"",""harness"":""Google DeepMind eval_hub / Context Arena"",""judge"":""hash check + SequenceMatcher strict scorer"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""none""}",True,official_model_card,verified,"Google Gemini 3.6 Flash July 2026 matrix: GDM MRCR v2 8-needle: 128k average. Exact effort, tools, sampling and harness are preserved."
grok-4.5,gdm_mrcr_v2_8needle_upto_128k,81.4,https://deepmind.google/models/model-cards/gemini-3-6-flash/,"{""effort"":""maximum/best available"",""harness"":""Google DeepMind eval_hub / Context Arena"",""judge"":""hash check + SequenceMatcher strict scorer"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""none""}",True,official_model_card,verified,"Google Gemini 3.6 Flash July 2026 matrix: GDM MRCR v2 8-needle: 128k average. Exact effort, tools, sampling and harness are preserved."
claude-sonnet-5,gdm_mrcr_v2_8needle_upto_128k,71.6,https://deepmind.google/models/model-cards/gemini-3-6-flash/,"{""effort"":""maximum/best available"",""harness"":""Google DeepMind eval_hub / Context Arena"",""judge"":""hash check + SequenceMatcher strict scorer"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""none""}",True,official_model_card,verified,"Google Gemini 3.6 Flash July 2026 matrix: GDM MRCR v2 8-needle: 128k average. Exact effort, tools, sampling and harness are preserved."
gemini-3.6-flash,gdm_mrcr_v2_8needle_1m,54.0,https://deepmind.google/models/model-cards/gemini-3-6-flash/,"{""effort"":""medium (default)"",""harness"":""Google DeepMind eval_hub / Context Arena"",""judge"":""hash check + SequenceMatcher strict scorer"",""mode"":""thinking/reasoning"",""sampling"":""pass@1"",""snapshot"":""July 2026"",""temperature"":""default"",""tools"":""none""}",True,official_model_card,verified,"Google Gemini 3.6 Flash July 2026 matrix: GDM MRCR v2 8-needle: 1M pointwise. Exact effort, tools, sampling and harness are preserved."
deepseek-v3.2-exp-thinking,text_arena_elo,1425.0,https://deepmind.google/models/gemma/gemma-4/,"{""context"":""default"",""effort"":""unknown"",""harness"":""Arena Text"",""judge"":""blind human raters"",""mode"":""thinking"",""prompt_style"":""live user prompts"",""sampling"":""live human pairwise votes"",""snapshot_date"":""2026-04-02"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,DeepMind page accessibility text provides the exact plotted Elo value.
deepseek-v4-flash,text_arena_elo,1436.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""unknown"",""harness"":""Arena Text"",""judge"":""blind human raters"",""mode"":""thinking"",""prompt_style"":""live user prompts"",""sampling"":""live human pairwise votes"",""snapshot_date"":""2026-06-19"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 4 dated Arena Text Elo snapshot.
deepseek-v4-pro,text_arena_elo,1458.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""unknown"",""harness"":""Arena Text"",""judge"":""blind human raters"",""mode"":""thinking"",""prompt_style"":""live user prompts"",""sampling"":""live human pairwise votes"",""snapshot_date"":""2026-06-19"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 4 dated Arena Text Elo snapshot.
deepseek-v4-pro-non-thinking,text_arena_elo,1456.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""unknown"",""harness"":""Arena Text"",""judge"":""blind human raters"",""mode"":""non-thinking"",""prompt_style"":""live user prompts"",""sampling"":""live human pairwise votes"",""snapshot_date"":""2026-06-19"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 4 dated Arena Text Elo snapshot.
gemma-3-27b,aime_2026,20.8,https://blog.google/innovation-and-ai/technology/developers-tools/gemma-4/,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,official_blog,verified,"Google Gemma 4 family launch table, duplicated by the DeepMind semantic table."
gemma-3-27b,codeforces_rating,110.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""code execution in undisclosed harness""}",False,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-3-27b,gdm_mrcr_v2_8needle_upto_128k,13.5,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-3-27b,graphwalks_lt128k_combined,32.8,https://arxiv.org/abs/2607.02770,"{""context"":""below 128K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-3-27b,ifbench,32.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-3-27b,infographicvqa,70.6,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none"",""visual_tokens"":""Pan & Scan""}",True,tech_report,verified,"Gemma 4 Technical Report Table 6; maximum supported resolution, 1,120 visual tokens."
gemma-3-27b,livecodebench_v6,29.1,https://blog.google/innovation-and-ai/technology/developers-tools/gemma-4/,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,official_blog,verified,"Google Gemma 4 family launch table, duplicated by the DeepMind semantic table."
gemma-3-27b,loft_text_retrieval_128k,8.6,https://arxiv.org/abs/2607.02770,"{""context"":""128K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-3-27b,mathvision,46.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none"",""visual_tokens"":""Pan & Scan""}",True,tech_report,verified,"Gemma 4 Technical Report Table 6; maximum supported resolution, 1,120 visual tokens."
gemma-3-27b,mmmlu,70.7,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-3-27b,mmmu_pro,49.7,https://blog.google/innovation-and-ai/technology/developers-tools/gemma-4/,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none"",""visual_tokens"":""Pan & Scan""}",True,official_blog,verified,"Google Gemma 4 family launch table, duplicated by the DeepMind semantic table."
gemma-3-27b,mtob_eng_kgv_half_book,41.0,https://arxiv.org/abs/2607.02770,"{""context"":""about 128K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-3-27b,mtob_kgv_eng_half_book,31.2,https://arxiv.org/abs/2607.02770,"{""context"":""about 128K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-3-27b,omnidocbench_1.5,0.365,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none"",""visual_tokens"":""Pan & Scan""}",True,tech_report,verified,"Gemma 4 Technical Report Table 6; maximum supported resolution, 1,120 visual tokens."
gemma-3-27b,ruler_128k,66.0,https://arxiv.org/abs/2607.02770,"{""context"":""128K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-3-27b,ruler_32k,91.1,https://arxiv.org/abs/2607.02770,"{""context"":""32K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-3-27b,scicode,21.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""unit tests"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-3-27b,tau2_bench_airline,39.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""Tau2-bench"",""judge"":""state-based task success"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""benchmark APIs + simulated user""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-3-27b,tau2_bench_avg,16.2,https://ai.google.dev/gemma/docs/core/model_card_4,"{""context"":""default"",""effort"":""n/a"",""harness"":""Tau2-bench"",""judge"":""state-based task success"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""benchmark APIs + simulated user""}",True,model_card,verified,Google model card matrix; all values are instruction-tuned despite the duplicated base-card surface.
gemma-3-27b,tau2_bench_retail,6.6,https://blog.google/innovation-and-ai/technology/developers-tools/gemma-4/,"{""context"":""default"",""effort"":""n/a"",""harness"":""Tau2-bench"",""judge"":""state-based task success"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""benchmark APIs + simulated user""}",True,official_blog,verified,"Google Gemma 4 family launch table, duplicated by the DeepMind semantic table."
gemma-3-27b,tau2_bench_telecom,3.1,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""Tau2-bench"",""judge"":""state-based task success"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""benchmark APIs + simulated user""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-3-27b,terminal_bench_hard,4.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""Terminal-Bench Hard provider harness"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""agentic terminal""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-3-27b,text_arena_elo,1365.0,https://blog.google/innovation-and-ai/technology/developers-tools/gemma-4/,"{""context"":""default"",""effort"":""n/a"",""harness"":""Arena AI text"",""judge"":""blind human raters"",""mode"":""non-thinking"",""prompt_style"":""live user prompts"",""sampling"":""live human pairwise votes"",""snapshot_date"":""2026-04-02"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,"Google Gemma 4 family launch table, duplicated by the DeepMind semantic table."
gemma-3n-e2b-instructed,covost2_de_en,36.5,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""CorpusBLEU"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7, transcribe-then-translate CorpusBLEU."
gemma-3n-e2b-instructed,covost2_es_en,39.9,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""CorpusBLEU"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7, transcribe-then-translate CorpusBLEU."
gemma-3n-e2b-instructed,covost2_fr_en,35.7,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""CorpusBLEU"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7, transcribe-then-translate CorpusBLEU."
gemma-3n-e2b-instructed,covost2_it_en,38.5,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""CorpusBLEU"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7, transcribe-then-translate CorpusBLEU."
gemma-3n-e2b-instructed,covost2_ja_en,17.7,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""CorpusBLEU"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7, transcribe-then-translate CorpusBLEU."
gemma-3n-e2b-instructed,covost2_ru_en,39.2,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""CorpusBLEU"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7, transcribe-then-translate CorpusBLEU."
gemma-3n-e2b-instructed,covost2_xx_en_7lang_macro,31.6,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""CorpusBLEU"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7, transcribe-then-translate CorpusBLEU."
gemma-3n-e2b-instructed,covost2_zh_cn_en,13.9,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""CorpusBLEU"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7, transcribe-then-translate CorpusBLEU."
gemma-3n-e2b-instructed,fleurs,0.108,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-3n-e2b-instructed,fleurs_asr_ar_eg,0.131,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-3n-e2b-instructed,fleurs_asr_de_de,0.079,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-3n-e2b-instructed,fleurs_asr_en_us,0.076,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-3n-e2b-instructed,fleurs_asr_es_419,0.051,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-3n-e2b-instructed,fleurs_asr_fr_fr,0.13,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-3n-e2b-instructed,fleurs_asr_hi_in,0.106,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-3n-e2b-instructed,fleurs_asr_it_it,0.044,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-3n-e2b-instructed,fleurs_asr_ja_jp_cer,0.163,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-3n-e2b-instructed,fleurs_asr_ko_kr_cer,0.101,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-3n-e2b-instructed,fleurs_asr_pt_br,0.067,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-3n-e2b-instructed,fleurs_asr_ru_ru,0.112,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-3n-e2b-instructed,fleurs_asr_zh_cn_cer,0.235,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-3n-e4b-instructed,covost2_de_en,39.1,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""CorpusBLEU"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7, transcribe-then-translate CorpusBLEU."
gemma-3n-e4b-instructed,covost2_es_en,41.8,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""CorpusBLEU"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7, transcribe-then-translate CorpusBLEU."
gemma-3n-e4b-instructed,covost2_fr_en,38.4,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""CorpusBLEU"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7, transcribe-then-translate CorpusBLEU."
gemma-3n-e4b-instructed,covost2_it_en,40.4,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""CorpusBLEU"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7, transcribe-then-translate CorpusBLEU."
gemma-3n-e4b-instructed,covost2_ja_en,22.3,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""CorpusBLEU"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7, transcribe-then-translate CorpusBLEU."
gemma-3n-e4b-instructed,covost2_ru_en,43.7,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""CorpusBLEU"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7, transcribe-then-translate CorpusBLEU."
gemma-3n-e4b-instructed,covost2_xx_en_7lang_macro,34.7,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""CorpusBLEU"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7, transcribe-then-translate CorpusBLEU."
gemma-3n-e4b-instructed,covost2_zh_cn_en,17.4,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""CorpusBLEU"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7, transcribe-then-translate CorpusBLEU."
gemma-3n-e4b-instructed,fleurs,0.085,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-3n-e4b-instructed,fleurs_asr_ar_eg,0.101,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-3n-e4b-instructed,fleurs_asr_de_de,0.065,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-3n-e4b-instructed,fleurs_asr_en_us,0.066,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-3n-e4b-instructed,fleurs_asr_es_419,0.041,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-3n-e4b-instructed,fleurs_asr_fr_fr,0.098,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-3n-e4b-instructed,fleurs_asr_hi_in,0.089,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-3n-e4b-instructed,fleurs_asr_it_it,0.034,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-3n-e4b-instructed,fleurs_asr_ja_jp_cer,0.111,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-3n-e4b-instructed,fleurs_asr_ko_kr_cer,0.073,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-3n-e4b-instructed,fleurs_asr_pt_br,0.053,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-3n-e4b-instructed,fleurs_asr_ru_ru,0.087,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-3n-e4b-instructed,fleurs_asr_zh_cn_cer,0.203,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-4-12b,aime_2026,77.5,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-12b,bigbench_extra_hard,53.0,https://blog.google/innovation-and-ai/technology/developers-tools/introducing-gemma-4-12B/,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,official_blog,verified,Google June 3 Gemma 4 12B launch raster; value is printed above the 12B bar in the original 1920x1080 asset.
gemma-4-12b,codeforces_rating,1659.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""code execution in undisclosed harness""}",False,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-12b,covost2,38.5,https://ai.google.dev/gemma/docs/core/model_card_4,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""language_scope"":""excluding Chinese"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,model_card,verified,Google model card matrix; all values are instruction-tuned despite the duplicated base-card surface.
gemma-4-12b,covost2_de_en,41.9,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""CorpusBLEU"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 8, encoder-free 12B audio evaluation."
gemma-4-12b,covost2_es_en,44.6,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""CorpusBLEU"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 8, encoder-free 12B audio evaluation."
gemma-4-12b,covost2_fr_en,42.5,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""CorpusBLEU"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 8, encoder-free 12B audio evaluation."
gemma-4-12b,covost2_it_en,43.3,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""CorpusBLEU"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 8, encoder-free 12B audio evaluation."
gemma-4-12b,covost2_ja_en,26.4,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""CorpusBLEU"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 8, encoder-free 12B audio evaluation."
gemma-4-12b,covost2_ru_en,50.5,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""CorpusBLEU"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 8, encoder-free 12B audio evaluation."
gemma-4-12b,docvqa,94.9,https://blog.google/innovation-and-ai/technology/developers-tools/introducing-gemma-4-12B/,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none"",""visual_tokens"":1120}",True,official_blog,verified,Google June 3 Gemma 4 12B launch raster; value is printed above the 12B bar in the original 1920x1080 asset.
gemma-4-12b,fleurs,0.069,https://ai.google.dev/gemma/docs/core/model_card_4,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""language_scope"":""excluding Chinese"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,model_card,verified,Google model card matrix; all values are instruction-tuned despite the duplicated base-card surface.
gemma-4-12b,fleurs_asr_ar_eg,0.07,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 8, encoder-free 12B audio evaluation."
gemma-4-12b,fleurs_asr_de_de,0.053,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 8, encoder-free 12B audio evaluation."
gemma-4-12b,fleurs_asr_en_us,0.063,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 8, encoder-free 12B audio evaluation."
gemma-4-12b,fleurs_asr_es_419,0.038,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 8, encoder-free 12B audio evaluation."
gemma-4-12b,fleurs_asr_fr_fr,0.081,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 8, encoder-free 12B audio evaluation."
gemma-4-12b,fleurs_asr_it_it,0.03,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 8, encoder-free 12B audio evaluation."
gemma-4-12b,fleurs_asr_ja_jp_cer,0.08,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 8, encoder-free 12B audio evaluation."
gemma-4-12b,fleurs_asr_ko_kr_cer,0.057,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 8, encoder-free 12B audio evaluation."
gemma-4-12b,fleurs_asr_pt_br,0.047,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 8, encoder-free 12B audio evaluation."
gemma-4-12b,fleurs_asr_ru_ru,0.068,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 8, encoder-free 12B audio evaluation."
gemma-4-12b,gdm_mrcr_v2_8needle_upto_128k,43.4,https://blog.google/innovation-and-ai/technology/developers-tools/introducing-gemma-4-12B/,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,official_blog,verified,Google June 3 Gemma 4 12B launch raster; value is printed above the 12B bar in the original 1920x1080 asset.
gemma-4-12b,gpqa_diamond,78.8,https://blog.google/innovation-and-ai/technology/developers-tools/introducing-gemma-4-12B/,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,official_blog,verified,Google June 3 Gemma 4 12B launch raster; value is printed above the 12B bar in the original 1920x1080 asset.
gemma-4-12b,graphwalks_lt128k_combined,71.0,https://arxiv.org/abs/2607.02770,"{""context"":""below 128K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-4-12b,hle,5.2,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-12b,ifbench,74.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-12b,ifeval,97.2,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-12b,infographicvqa,88.4,https://blog.google/innovation-and-ai/technology/developers-tools/introducing-gemma-4-12B/,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none"",""visual_tokens"":1120}",True,official_blog,verified,Google June 3 Gemma 4 12B launch raster; value is printed above the 12B bar in the original 1920x1080 asset.
gemma-4-12b,livecodebench_v6,72.0,https://blog.google/innovation-and-ai/technology/developers-tools/introducing-gemma-4-12B/,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,official_blog,verified,Google June 3 Gemma 4 12B launch raster; value is printed above the 12B bar in the original 1920x1080 asset.
gemma-4-12b,loft_text_retrieval_128k,66.4,https://arxiv.org/abs/2607.02770,"{""context"":""128K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-4-12b,mathvision,79.7,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none"",""visual_tokens"":1120}",True,tech_report,verified,"Gemma 4 Technical Report Table 6; maximum supported resolution, 1,120 visual tokens."
gemma-4-12b,medxpertqa_mm,48.7,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none"",""visual_tokens"":1120}",True,tech_report,verified,"Gemma 4 Technical Report Table 6; maximum supported resolution, 1,120 visual tokens."
gemma-4-12b,mmlu_pro,77.2,https://blog.google/innovation-and-ai/technology/developers-tools/introducing-gemma-4-12B/,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,official_blog,verified,Google June 3 Gemma 4 12B launch raster; value is printed above the 12B bar in the original 1920x1080 asset.
gemma-4-12b,mmmlu,83.4,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-12b,mmmu_pro,69.1,https://blog.google/innovation-and-ai/technology/developers-tools/introducing-gemma-4-12B/,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none"",""visual_tokens"":1120}",True,official_blog,verified,Google June 3 Gemma 4 12B launch raster; value is printed above the 12B bar in the original 1920x1080 asset.
gemma-4-12b,mtob_eng_kgv_full_book,41.9,https://arxiv.org/abs/2607.02770,"{""context"":""about 256K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-4-12b,mtob_eng_kgv_half_book,45.1,https://arxiv.org/abs/2607.02770,"{""context"":""about 128K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-4-12b,mtob_kgv_eng_full_book,32.9,https://arxiv.org/abs/2607.02770,"{""context"":""about 256K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-4-12b,mtob_kgv_eng_half_book,37.3,https://arxiv.org/abs/2607.02770,"{""context"":""about 128K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-4-12b,omnidocbench_1.5,0.164,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none"",""visual_tokens"":1120}",True,tech_report,verified,"Gemma 4 Technical Report Table 6; maximum supported resolution, 1,120 visual tokens."
gemma-4-12b,ruler_128k,91.2,https://arxiv.org/abs/2607.02770,"{""context"":""128K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-4-12b,ruler_32k,96.4,https://arxiv.org/abs/2607.02770,"{""context"":""32K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-4-12b,scicode,38.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""unit tests"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-12b,tau2_bench_airline,75.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""Tau2-bench"",""judge"":""state-based task success"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""benchmark APIs + simulated user""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-12b,tau2_bench_avg,69.0,https://ai.google.dev/gemma/docs/core/model_card_4,"{""context"":""default"",""effort"":""default"",""harness"":""Tau2-bench"",""judge"":""state-based task success"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""benchmark APIs + simulated user""}",True,model_card,verified,Google model card matrix; all values are instruction-tuned despite the duplicated base-card surface.
gemma-4-12b,tau2_bench_retail,77.6,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""Tau2-bench"",""judge"":""state-based task success"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""benchmark APIs + simulated user""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-12b,tau2_bench_telecom,54.4,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""Tau2-bench"",""judge"":""state-based task success"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""benchmark APIs + simulated user""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-12b,terminal_bench_hard,18.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""Terminal-Bench Hard provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""agentic terminal""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-26b-a4b,gdm_mrcr_v2_8needle_upto_128k,44.1,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-26b-a4b,graphwalks_lt128k_combined,72.6,https://arxiv.org/abs/2607.02770,"{""context"":""below 128K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-4-26b-a4b,ifbench,72.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-26b-a4b,ifeval,98.7,https://arxiv.org/abs/2608.00146,"{""context"":""default"",""decoding"":""AR MTP"",""harness"":""official Google cross-model evaluation"",""judge"":""IFEval strict/loose subtype undisclosed"",""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampling"":""replicate count N undisclosed"",""tools"":""none""}",False,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
gemma-4-26b-a4b,infographicvqa,89.3,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none"",""visual_tokens"":1120}",True,tech_report,verified,"Gemma 4 Technical Report Table 6; maximum supported resolution, 1,120 visual tokens."
gemma-4-26b-a4b,livecodebench_v6,77.1,https://blog.google/innovation-and-ai/technology/developers-tools/gemma-4/,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,official_blog,verified,"Google Gemma 4 family launch table, duplicated by the DeepMind semantic table."
gemma-4-26b-a4b,loft_text_retrieval_128k,66.3,https://arxiv.org/abs/2607.02770,"{""context"":""128K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-4-26b-a4b,mtob_eng_kgv_full_book,48.9,https://arxiv.org/abs/2607.02770,"{""context"":""about 256K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-4-26b-a4b,mtob_eng_kgv_half_book,50.0,https://arxiv.org/abs/2607.02770,"{""context"":""about 128K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-4-26b-a4b,mtob_kgv_eng_full_book,42.7,https://arxiv.org/abs/2607.02770,"{""context"":""about 256K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-4-26b-a4b,mtob_kgv_eng_half_book,45.0,https://arxiv.org/abs/2607.02770,"{""context"":""about 128K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-4-26b-a4b,omnidocbench_1.5,0.149,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none"",""visual_tokens"":1120}",True,tech_report,verified,"Gemma 4 Technical Report Table 6; maximum supported resolution, 1,120 visual tokens."
gemma-4-26b-a4b,ruler_128k,89.8,https://arxiv.org/abs/2607.02770,"{""context"":""128K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-4-26b-a4b,ruler_32k,97.3,https://arxiv.org/abs/2607.02770,"{""context"":""32K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-4-26b-a4b,scicode,40.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""unit tests"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-26b-a4b,tau2_bench_airline,76.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""Tau2-bench"",""judge"":""state-based task success"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""benchmark APIs + simulated user""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-26b-a4b,tau2_bench_avg,68.2,https://ai.google.dev/gemma/docs/core/model_card_4,"{""context"":""default"",""effort"":""default"",""harness"":""Tau2-bench"",""judge"":""state-based task success"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""benchmark APIs + simulated user""}",True,model_card,verified,Google model card matrix; all values are instruction-tuned despite the duplicated base-card surface.
gemma-4-26b-a4b,tau2_bench_retail,85.5,https://blog.google/innovation-and-ai/technology/developers-tools/gemma-4/,"{""context"":""default"",""effort"":""default"",""harness"":""Tau2-bench"",""judge"":""state-based task success"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""benchmark APIs + simulated user""}",True,official_blog,verified,"Google Gemma 4 family launch table, duplicated by the DeepMind semantic table."
gemma-4-26b-a4b,tau2_bench_telecom,43.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""Tau2-bench"",""judge"":""state-based task success"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""benchmark APIs + simulated user""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-26b-a4b,terminal_bench_hard,14.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""Terminal-Bench Hard provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""agentic terminal""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-26b-a4b,text_arena_elo,1441.0,https://blog.google/innovation-and-ai/technology/developers-tools/gemma-4/,"{""context"":""default"",""effort"":""default"",""harness"":""Arena AI text"",""judge"":""blind human raters"",""mode"":""thinking"",""prompt_style"":""live user prompts"",""sampling"":""live human pairwise votes"",""snapshot_date"":""2026-04-02"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,"Google Gemma 4 family launch table, duplicated by the DeepMind semantic table."
gemma-4-31b,gdm_mrcr_v2_8needle_upto_128k,66.4,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-31b,graphwalks_lt128k_combined,82.3,https://arxiv.org/abs/2607.02770,"{""context"":""below 128K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-4-31b,ifbench,76.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-31b,ifeval,98.9,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-31b,infographicvqa,92.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none"",""visual_tokens"":1120}",True,tech_report,verified,"Gemma 4 Technical Report Table 6; maximum supported resolution, 1,120 visual tokens."
gemma-4-31b,livecodebench_v6,80.0,https://blog.google/innovation-and-ai/technology/developers-tools/gemma-4/,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,official_blog,verified,"Google Gemma 4 family launch table, duplicated by the DeepMind semantic table."
gemma-4-31b,loft_text_retrieval_128k,79.5,https://arxiv.org/abs/2607.02770,"{""context"":""128K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-4-31b,mtob_eng_kgv_full_book,54.3,https://arxiv.org/abs/2607.02770,"{""context"":""about 256K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-4-31b,mtob_eng_kgv_half_book,52.9,https://arxiv.org/abs/2607.02770,"{""context"":""about 128K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-4-31b,mtob_kgv_eng_full_book,46.2,https://arxiv.org/abs/2607.02770,"{""context"":""about 256K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-4-31b,mtob_kgv_eng_half_book,48.6,https://arxiv.org/abs/2607.02770,"{""context"":""about 128K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-4-31b,omnidocbench_1.5,0.131,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none"",""visual_tokens"":1120}",True,tech_report,verified,"Gemma 4 Technical Report Table 6; maximum supported resolution, 1,120 visual tokens."
gemma-4-31b,ruler_128k,96.4,https://arxiv.org/abs/2607.02770,"{""context"":""128K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-4-31b,ruler_32k,96.8,https://arxiv.org/abs/2607.02770,"{""context"":""32K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-4-31b,scicode,43.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""unit tests"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-31b,tau2_bench_airline,75.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""Tau2-bench"",""judge"":""state-based task success"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""benchmark APIs + simulated user""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-31b,tau2_bench_avg,76.9,https://ai.google.dev/gemma/docs/core/model_card_4,"{""context"":""default"",""effort"":""default"",""harness"":""Tau2-bench"",""judge"":""state-based task success"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""benchmark APIs + simulated user""}",True,model_card,verified,Google model card matrix; all values are instruction-tuned despite the duplicated base-card surface.
gemma-4-31b,tau2_bench_retail,86.4,https://blog.google/innovation-and-ai/technology/developers-tools/gemma-4/,"{""context"":""default"",""effort"":""default"",""harness"":""Tau2-bench"",""judge"":""state-based task success"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""benchmark APIs + simulated user""}",True,official_blog,verified,"Google Gemma 4 family launch table, duplicated by the DeepMind semantic table."
gemma-4-31b,tau2_bench_telecom,69.3,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""Tau2-bench"",""judge"":""state-based task success"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""benchmark APIs + simulated user""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-31b,terminal_bench_hard,36.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""Terminal-Bench Hard provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""agentic terminal""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-31b,text_arena_elo,1452.0,https://blog.google/innovation-and-ai/technology/developers-tools/gemma-4/,"{""context"":""default"",""effort"":""default"",""harness"":""Arena AI text"",""judge"":""blind human raters"",""mode"":""thinking"",""prompt_style"":""live user prompts"",""sampling"":""live human pairwise votes"",""snapshot_date"":""2026-04-02"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,"Google Gemma 4 family launch table, duplicated by the DeepMind semantic table."
gemma-4-e2b,covost2,33.47,https://ai.google.dev/gemma/docs/core/model_card_4,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""language_scope"":""provider-reported aggregate"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,model_card,verified,Google model card matrix; all values are instruction-tuned despite the duplicated base-card surface.
gemma-4-e2b,covost2_de_en,39.2,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""CorpusBLEU"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7, transcribe-then-translate CorpusBLEU."
gemma-4-e2b,covost2_es_en,43.2,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""CorpusBLEU"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7, transcribe-then-translate CorpusBLEU."
gemma-4-e2b,covost2_fr_en,39.2,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""CorpusBLEU"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7, transcribe-then-translate CorpusBLEU."
gemma-4-e2b,covost2_it_en,40.8,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""CorpusBLEU"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7, transcribe-then-translate CorpusBLEU."
gemma-4-e2b,covost2_ja_en,21.4,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""CorpusBLEU"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7, transcribe-then-translate CorpusBLEU."
gemma-4-e2b,covost2_ru_en,46.4,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""CorpusBLEU"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7, transcribe-then-translate CorpusBLEU."
gemma-4-e2b,covost2_xx_en_7lang_macro,35.4,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""CorpusBLEU"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7, transcribe-then-translate CorpusBLEU."
gemma-4-e2b,covost2_zh_cn_en,17.9,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""CorpusBLEU"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7, transcribe-then-translate CorpusBLEU."
gemma-4-e2b,fleurs_asr_ar_eg,0.143,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-4-e2b,fleurs_asr_de_de,0.076,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-4-e2b,fleurs_asr_en_us,0.08,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-4-e2b,fleurs_asr_es_419,0.042,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-4-e2b,fleurs_asr_fr_fr,0.101,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-4-e2b,fleurs_asr_hi_in,0.101,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-4-e2b,fleurs_asr_it_it,0.041,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-4-e2b,fleurs_asr_ja_jp_cer,0.107,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-4-e2b,fleurs_asr_ko_kr_cer,0.066,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-4-e2b,fleurs_asr_pt_br,0.056,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-4-e2b,fleurs_asr_ru_ru,0.084,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-4-e2b,fleurs_asr_zh_cn_cer,0.187,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-4-e2b,gdm_mrcr_v2_8needle_upto_128k,19.1,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-e2b,graphwalks_lt128k_combined,4.1,https://arxiv.org/abs/2607.02770,"{""context"":""below 128K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-4-e2b,ifbench,38.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-e2b,ifeval,94.6,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-e2b,infographicvqa,63.9,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none"",""visual_tokens"":1120}",True,tech_report,verified,"Gemma 4 Technical Report Table 6; maximum supported resolution, 1,120 visual tokens."
gemma-4-e2b,livecodebench_v6,44.0,https://blog.google/innovation-and-ai/technology/developers-tools/gemma-4/,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,official_blog,verified,"Google Gemma 4 family launch table, duplicated by the DeepMind semantic table."
gemma-4-e2b,loft_text_retrieval_128k,50.5,https://arxiv.org/abs/2607.02770,"{""context"":""128K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-4-e2b,mtob_eng_kgv_half_book,15.4,https://arxiv.org/abs/2607.02770,"{""context"":""about 128K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-4-e2b,mtob_kgv_eng_half_book,28.2,https://arxiv.org/abs/2607.02770,"{""context"":""about 128K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-4-e2b,omnidocbench_1.5,0.29,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none"",""visual_tokens"":1120}",True,tech_report,verified,"Gemma 4 Technical Report Table 6; maximum supported resolution, 1,120 visual tokens."
gemma-4-e2b,ruler_128k,70.4,https://arxiv.org/abs/2607.02770,"{""context"":""128K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-4-e2b,ruler_32k,83.0,https://arxiv.org/abs/2607.02770,"{""context"":""32K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-4-e2b,scicode,21.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""unit tests"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-e2b,tau2_bench_airline,31.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""Tau2-bench"",""judge"":""state-based task success"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""benchmark APIs + simulated user""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-e2b,tau2_bench_avg,24.5,https://ai.google.dev/gemma/docs/core/model_card_4,"{""context"":""default"",""effort"":""default"",""harness"":""Tau2-bench"",""judge"":""state-based task success"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""benchmark APIs + simulated user""}",True,model_card,verified,Google model card matrix; all values are instruction-tuned despite the duplicated base-card surface.
gemma-4-e2b,tau2_bench_retail,29.4,https://blog.google/innovation-and-ai/technology/developers-tools/gemma-4/,"{""context"":""default"",""effort"":""default"",""harness"":""Tau2-bench"",""judge"":""state-based task success"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""benchmark APIs + simulated user""}",True,official_blog,verified,"Google Gemma 4 family launch table, duplicated by the DeepMind semantic table."
gemma-4-e2b,tau2_bench_telecom,19.7,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""Tau2-bench"",""judge"":""state-based task success"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""benchmark APIs + simulated user""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-e2b,terminal_bench_hard,3.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""Terminal-Bench Hard provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""agentic terminal""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-e4b,covost2,35.54,https://ai.google.dev/gemma/docs/core/model_card_4,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""language_scope"":""provider-reported aggregate"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,model_card,verified,Google model card matrix; all values are instruction-tuned despite the duplicated base-card surface.
gemma-4-e4b,covost2_de_en,42.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""CorpusBLEU"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7, transcribe-then-translate CorpusBLEU."
gemma-4-e4b,covost2_es_en,44.8,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""CorpusBLEU"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7, transcribe-then-translate CorpusBLEU."
gemma-4-e4b,covost2_fr_en,41.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""CorpusBLEU"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7, transcribe-then-translate CorpusBLEU."
gemma-4-e4b,covost2_it_en,43.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""CorpusBLEU"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7, transcribe-then-translate CorpusBLEU."
gemma-4-e4b,covost2_ja_en,25.5,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""CorpusBLEU"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7, transcribe-then-translate CorpusBLEU."
gemma-4-e4b,covost2_ru_en,49.4,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""CorpusBLEU"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7, transcribe-then-translate CorpusBLEU."
gemma-4-e4b,covost2_xx_en_7lang_macro,38.2,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""CorpusBLEU"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7, transcribe-then-translate CorpusBLEU."
gemma-4-e4b,covost2_zh_cn_en,21.9,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""CorpusBLEU"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7, transcribe-then-translate CorpusBLEU."
gemma-4-e4b,fleurs_asr_ar_eg,0.162,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-4-e4b,fleurs_asr_de_de,0.061,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-4-e4b,fleurs_asr_en_us,0.065,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-4-e4b,fleurs_asr_es_419,0.035,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-4-e4b,fleurs_asr_fr_fr,0.08,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-4-e4b,fleurs_asr_hi_in,0.086,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-4-e4b,fleurs_asr_it_it,0.032,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-4-e4b,fleurs_asr_ja_jp_cer,0.078,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-4-e4b,fleurs_asr_ko_kr_cer,0.053,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-4-e4b,fleurs_asr_pt_br,0.046,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-4-e4b,fleurs_asr_ru_ru,0.068,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-4-e4b,fleurs_asr_zh_cn_cer,0.136,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""n/a"",""harness"":""official Google audio evaluation"",""judge"":""WER/CER"",""mode"":""unknown"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""audio input; no external tools""}",False,tech_report,verified,"Gemma 4 Technical Report Table 7; WER except CER for Korean, Japanese and Chinese."
gemma-4-e4b,gdm_mrcr_v2_8needle_upto_128k,25.4,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-e4b,graphwalks_lt128k_combined,50.9,https://arxiv.org/abs/2607.02770,"{""context"":""below 128K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-4-e4b,ifbench,44.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-e4b,ifeval,96.7,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-e4b,infographicvqa,70.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none"",""visual_tokens"":1120}",True,tech_report,verified,"Gemma 4 Technical Report Table 6; maximum supported resolution, 1,120 visual tokens."
gemma-4-e4b,livecodebench_v6,52.0,https://blog.google/innovation-and-ai/technology/developers-tools/gemma-4/,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,official_blog,verified,"Google Gemma 4 family launch table, duplicated by the DeepMind semantic table."
gemma-4-e4b,loft_text_retrieval_128k,58.5,https://arxiv.org/abs/2607.02770,"{""context"":""128K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-4-e4b,mtob_eng_kgv_half_book,37.8,https://arxiv.org/abs/2607.02770,"{""context"":""about 128K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-4-e4b,mtob_kgv_eng_half_book,34.6,https://arxiv.org/abs/2607.02770,"{""context"":""about 128K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-4-e4b,omnidocbench_1.5,0.181,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none"",""visual_tokens"":1120}",True,tech_report,verified,"Gemma 4 Technical Report Table 6; maximum supported resolution, 1,120 visual tokens."
gemma-4-e4b,ruler_128k,86.6,https://arxiv.org/abs/2607.02770,"{""context"":""128K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-4-e4b,ruler_32k,95.2,https://arxiv.org/abs/2607.02770,"{""context"":""32K"",""effort"":""n/a"",""harness"":""official Google evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 9; final IT models evaluated without thinking.
gemma-4-e4b,scicode,24.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""official Google evaluation"",""judge"":""unit tests"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""none""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-e4b,tau2_bench_airline,52.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""Tau2-bench"",""judge"":""state-based task success"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""benchmark APIs + simulated user""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-e4b,tau2_bench_avg,42.2,https://ai.google.dev/gemma/docs/core/model_card_4,"{""context"":""default"",""effort"":""default"",""harness"":""Tau2-bench"",""judge"":""state-based task success"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""benchmark APIs + simulated user""}",True,model_card,verified,Google model card matrix; all values are instruction-tuned despite the duplicated base-card surface.
gemma-4-e4b,tau2_bench_retail,57.5,https://blog.google/innovation-and-ai/technology/developers-tools/gemma-4/,"{""context"":""default"",""effort"":""default"",""harness"":""Tau2-bench"",""judge"":""state-based task success"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""benchmark APIs + simulated user""}",True,official_blog,verified,"Google Gemma 4 family launch table, duplicated by the DeepMind semantic table."
gemma-4-e4b,tau2_bench_telecom,18.4,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""Tau2-bench"",""judge"":""state-based task success"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""benchmark APIs + simulated user""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
gemma-4-e4b,terminal_bench_hard,8.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""default"",""harness"":""Terminal-Bench Hard provider harness"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""provider-reported"",""temperature"":""not disclosed"",""tools"":""agentic terminal""}",True,tech_report,verified,"Gemma 4 Technical Report Table 5; Gemma 4 thinking unless explicitly stated, Gemma 3 27B non-thinking."
glm-5,text_arena_elo,1456.0,https://deepmind.google/models/gemma/gemma-4/,"{""context"":""default"",""effort"":""unknown"",""harness"":""Arena Text"",""judge"":""blind human raters"",""mode"":""unknown"",""prompt_style"":""live user prompts"",""sampling"":""live human pairwise votes"",""snapshot_date"":""2026-04-02"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,DeepMind page accessibility text provides the exact plotted Elo value.
glm-5.1,text_arena_elo,1475.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""unknown"",""harness"":""Arena Text"",""judge"":""blind human raters"",""mode"":""unknown"",""prompt_style"":""live user prompts"",""sampling"":""live human pairwise votes"",""snapshot_date"":""2026-06-19"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 4 dated Arena Text Elo snapshot.
gpt-oss-120b,text_arena_elo,1354.0,https://deepmind.google/models/gemma/gemma-4/,"{""context"":""default"",""effort"":""unknown"",""harness"":""Arena Text"",""judge"":""blind human raters"",""mode"":""unknown"",""prompt_style"":""live user prompts"",""sampling"":""live human pairwise votes"",""snapshot_date"":""2026-04-02"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,DeepMind page accessibility text provides the exact plotted Elo value.
kimi-k2.5,text_arena_elo,1455.0,https://deepmind.google/models/gemma/gemma-4/,"{""context"":""default"",""effort"":""unknown"",""harness"":""Arena Text"",""judge"":""blind human raters"",""mode"":""thinking"",""prompt_style"":""live user prompts"",""sampling"":""live human pairwise votes"",""snapshot_date"":""2026-04-02"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,DeepMind page accessibility text provides the exact plotted Elo value.
kimi-k2.6,text_arena_elo,1460.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""unknown"",""harness"":""Arena Text"",""judge"":""blind human raters"",""mode"":""unknown"",""prompt_style"":""live user prompts"",""sampling"":""live human pairwise votes"",""snapshot_date"":""2026-06-19"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 4 dated Arena Text Elo snapshot.
mimo-v2.5-pro,text_arena_elo,1466.0,https://arxiv.org/abs/2607.02770,"{""context"":""default"",""effort"":""unknown"",""harness"":""Arena Text"",""judge"":""blind human raters"",""mode"":""unknown"",""prompt_style"":""live user prompts"",""sampling"":""live human pairwise votes"",""snapshot_date"":""2026-06-19"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Gemma 4 Technical Report Table 4 dated Arena Text Elo snapshot.
mistral-large-3,text_arena_elo,1416.0,https://deepmind.google/models/gemma/gemma-4/,"{""context"":""default"",""effort"":""unknown"",""harness"":""Arena Text"",""judge"":""blind human raters"",""mode"":""unknown"",""prompt_style"":""live user prompts"",""sampling"":""live human pairwise votes"",""snapshot_date"":""2026-04-02"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,DeepMind page accessibility text provides the exact plotted Elo value.
qwen3.5-122b-a10b,text_arena_elo,1416.0,https://deepmind.google/models/gemma/gemma-4/,"{""context"":""default"",""effort"":""unknown"",""harness"":""Arena Text"",""judge"":""blind human raters"",""mode"":""unknown"",""prompt_style"":""live user prompts"",""sampling"":""live human pairwise votes"",""snapshot_date"":""2026-04-02"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,DeepMind page accessibility text provides the exact plotted Elo value.
qwen3.5-27b,text_arena_elo,1404.0,https://deepmind.google/models/gemma/gemma-4/,"{""context"":""default"",""effort"":""unknown"",""harness"":""Arena Text"",""judge"":""blind human raters"",""mode"":""unknown"",""prompt_style"":""live user prompts"",""sampling"":""live human pairwise votes"",""snapshot_date"":""2026-04-02"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,DeepMind page accessibility text provides the exact plotted Elo value.
qwen3.5-397b,text_arena_elo,1450.0,https://deepmind.google/models/gemma/gemma-4/,"{""context"":""default"",""effort"":""unknown"",""harness"":""Arena Text"",""judge"":""blind human raters"",""mode"":""unknown"",""prompt_style"":""live user prompts"",""sampling"":""live human pairwise votes"",""snapshot_date"":""2026-04-02"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,DeepMind page accessibility text provides the exact plotted Elo value.
diffusiongemma-26b-a4b,aime_2026,69.1,https://blog.google/innovation-and-ai/technology/developers-tools/diffusion-gemma-faster-text-generation/,"{""canvas"":256,""context"":""default"",""decoding"":""TD"",""harness"":""official Google launch evaluation"",""judge"":""benchmark-specified"",""max_denoising_steps"":48,""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampler"":""entropy-bounded adaptive"",""sampling"":""replicate count N undisclosed"",""tools"":""none""}",True,official_blog,verified,Google release-blog benchmark chart; highest-priority duplicate for target quality values.
diffusiongemma-26b-a4b,bigbench_extra_hard,47.6,https://arxiv.org/abs/2608.00146,"{""canvas"":256,""context"":""default"",""decoding"":""TD"",""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""max_denoising_steps"":48,""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampler"":""entropy-bounded adaptive"",""sampling"":""replicate count N undisclosed"",""tools"":""none""}",True,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
diffusiongemma-26b-a4b,codeelo_rating,1429.0,https://arxiv.org/abs/2608.00146,"{""canvas"":256,""context"":""default"",""decoding"":""TD"",""harness"":""official Google cross-model evaluation"",""judge"":""Codeforces online judge"",""max_denoising_steps"":48,""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampler"":""entropy-bounded adaptive"",""sampling"":""Google run count undisclosed; canonical samples=8"",""tools"":""none""}",False,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
diffusiongemma-26b-a4b,gdm_mrcr_v2_8needle_upto_128k,32.0,https://huggingface.co/google/diffusiongemma-26B-A4B-it,"{""canvas"":256,""context"":""up to 128K"",""decoding"":""TD"",""harness"":""official Google model-card evaluation"",""judge"":""official MRCR strict scorer"",""max_denoising_steps"":48,""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampler"":""entropy-bounded adaptive"",""sampling"":""replicate count N undisclosed"",""tools"":""none""}",True,model_card,verified,Official HF model-card matrix; Kaggle is a duplicate physical mirror.
diffusiongemma-26b-a4b,gpqa_diamond,73.2,https://blog.google/innovation-and-ai/technology/developers-tools/diffusion-gemma-faster-text-generation/,"{""canvas"":256,""context"":""default"",""decoding"":""TD"",""harness"":""official Google launch evaluation"",""judge"":""benchmark-specified"",""max_denoising_steps"":48,""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampler"":""entropy-bounded adaptive"",""sampling"":""replicate count N undisclosed"",""tools"":""none""}",True,official_blog,verified,Google release-blog benchmark chart; highest-priority duplicate for target quality values.
diffusiongemma-26b-a4b,gsm8k,96.3,https://arxiv.org/abs/2608.00146,"{""canvas"":256,""context"":""default"",""decoding"":""TD"",""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""max_denoising_steps"":48,""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampler"":""entropy-bounded adaptive"",""sampling"":""replicate count N undisclosed"",""tools"":""none""}",True,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
diffusiongemma-26b-a4b,hiddenmath,80.6,https://arxiv.org/abs/2608.00146,"{""canvas"":256,""context"":""default"",""decoding"":""TD"",""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""max_denoising_steps"":48,""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampler"":""entropy-bounded adaptive"",""sampling"":""replicate count N undisclosed"",""tools"":""none""}",True,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
diffusiongemma-26b-a4b,hle,11.0,https://huggingface.co/google/diffusiongemma-26B-A4B-it,"{""canvas"":256,""context"":""default"",""decoding"":""TD"",""harness"":""official Google model-card evaluation"",""judge"":""HLE official rubric"",""max_denoising_steps"":48,""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampler"":""entropy-bounded adaptive"",""sampling"":""replicate count N undisclosed"",""tools"":""none""}",True,model_card,verified,Official HF model-card matrix; Kaggle is a duplicate physical mirror.
diffusiongemma-26b-a4b,humaneval,94.5,https://arxiv.org/abs/2608.00146,"{""canvas"":256,""context"":""default"",""decoding"":""TD"",""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""max_denoising_steps"":48,""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampler"":""entropy-bounded adaptive"",""sampling"":""replicate count N undisclosed"",""tools"":""none""}",True,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
diffusiongemma-26b-a4b,ifeval,97.4,https://arxiv.org/abs/2608.00146,"{""canvas"":256,""context"":""default"",""decoding"":""TD"",""harness"":""official Google cross-model evaluation"",""judge"":""IFEval strict/loose subtype undisclosed"",""max_denoising_steps"":48,""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampler"":""entropy-bounded adaptive"",""sampling"":""replicate count N undisclosed"",""tools"":""none""}",False,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
diffusiongemma-26b-a4b,lbpp_v2_multilingual,81.0,https://arxiv.org/abs/2608.00146,"{""canvas"":256,""context"":""default"",""decoding"":""TD"",""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""max_denoising_steps"":48,""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampler"":""entropy-bounded adaptive"",""sampling"":""replicate count N undisclosed"",""tools"":""none""}",True,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
diffusiongemma-26b-a4b,livecodebench_v6,69.1,https://blog.google/innovation-and-ai/technology/developers-tools/diffusion-gemma-faster-text-generation/,"{""canvas"":256,""context"":""default"",""decoding"":""TD"",""harness"":""official Google launch evaluation"",""judge"":""benchmark-specified"",""max_denoising_steps"":48,""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampler"":""entropy-bounded adaptive"",""sampling"":""replicate count N undisclosed"",""tools"":""none""}",True,official_blog,verified,Google release-blog benchmark chart; highest-priority duplicate for target quality values.
diffusiongemma-26b-a4b,mathvision,70.5,https://huggingface.co/google/diffusiongemma-26B-A4B-it,"{""canvas"":256,""context"":""default"",""decoding"":""TD"",""harness"":""official Google model-card evaluation"",""judge"":""benchmark-specified"",""max_denoising_steps"":48,""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampler"":""entropy-bounded adaptive"",""sampling"":""replicate count N undisclosed"",""tools"":""none"",""visual_tokens"":""undisclosed""}",False,model_card,verified,Official HF model-card matrix; Kaggle is a duplicate physical mirror.
diffusiongemma-26b-a4b,medxpertqa_mm,49.0,https://huggingface.co/google/diffusiongemma-26B-A4B-it,"{""canvas"":256,""context"":""default"",""decoding"":""TD"",""harness"":""official Google model-card evaluation"",""judge"":""benchmark-specified"",""max_denoising_steps"":48,""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampler"":""entropy-bounded adaptive"",""sampling"":""replicate count N undisclosed"",""tools"":""none"",""visual_tokens"":""undisclosed""}",False,model_card,verified,Official HF model-card matrix; Kaggle is a duplicate physical mirror.
diffusiongemma-26b-a4b,mgsm,84.8,https://arxiv.org/abs/2608.00146,"{""canvas"":256,""context"":""default"",""decoding"":""TD"",""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""max_denoising_steps"":48,""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampler"":""entropy-bounded adaptive"",""sampling"":""replicate count N undisclosed"",""tools"":""none""}",True,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
diffusiongemma-26b-a4b,mmlu_pro,77.6,https://blog.google/innovation-and-ai/technology/developers-tools/diffusion-gemma-faster-text-generation/,"{""canvas"":256,""context"":""default"",""decoding"":""TD"",""harness"":""official Google launch evaluation"",""judge"":""benchmark-specified"",""max_denoising_steps"":48,""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampler"":""entropy-bounded adaptive"",""sampling"":""replicate count N undisclosed"",""tools"":""none""}",True,official_blog,verified,Google release-blog benchmark chart; highest-priority duplicate for target quality values.
diffusiongemma-26b-a4b,mmmlu,81.5,https://blog.google/innovation-and-ai/technology/developers-tools/diffusion-gemma-faster-text-generation/,"{""canvas"":256,""context"":""default"",""decoding"":""TD"",""harness"":""official Google launch evaluation"",""judge"":""benchmark-specified"",""max_denoising_steps"":48,""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampler"":""entropy-bounded adaptive"",""sampling"":""replicate count N undisclosed"",""tools"":""none""}",True,official_blog,verified,Google release-blog benchmark chart; highest-priority duplicate for target quality values.
diffusiongemma-26b-a4b,mmmu_pro,54.3,https://arxiv.org/abs/2608.00146,"{""canvas"":256,""context"":""default"",""decoding"":""TD"",""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""max_denoising_steps"":48,""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampler"":""entropy-bounded adaptive"",""sampling"":""replicate count N undisclosed"",""tools"":""none"",""visual_tokens"":""undisclosed""}",False,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
diffusiongemma-26b-a4b,natural2code,94.0,https://arxiv.org/abs/2608.00146,"{""canvas"":256,""context"":""default"",""decoding"":""TD"",""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""max_denoising_steps"":48,""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampler"":""entropy-bounded adaptive"",""sampling"":""replicate count N undisclosed"",""tools"":""none""}",True,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
diffusiongemma-26b-a4b,omnidocbench_1.5,0.319,https://huggingface.co/google/diffusiongemma-26B-A4B-it,"{""canvas"":256,""context"":""default"",""decoding"":""TD"",""harness"":""official Google model-card evaluation"",""judge"":""benchmark-specified"",""max_denoising_steps"":48,""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampler"":""entropy-bounded adaptive"",""sampling"":""replicate count N undisclosed"",""tools"":""none"",""visual_tokens"":""undisclosed""}",False,model_card,verified,Official HF model-card matrix; Kaggle is a duplicate physical mirror.
diffusiongemma-26b-a4b,pubmedqa_pqal_decision_accuracy,75.6,https://arxiv.org/abs/2608.00146,"{""canvas"":128,""context"":""1024-token prompt"",""decoding"":""TD"",""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""max_denoising_steps"":48,""mode"":""thinking"",""number_of_canvases"":2,""prompt_style"":""official provider chat template"",""sampler"":""entropy-bounded adaptive"",""sampling"":""replicate count N undisclosed"",""tools"":""none""}",False,tech_report,verified,"Released DiffusionGemma before downstream LoRA. TD-thinking is inherited globally, but the PubMedQA recipe uses a noncanonical 128-token canvas with two canvases."
diffusiongemma-26b-a4b,pubmedqa_pqal_long_answer_bleu,10.76,https://arxiv.org/abs/2608.00146,"{""canvas"":128,""context"":""1024-token prompt"",""decoding"":""TD"",""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""max_denoising_steps"":48,""mode"":""thinking"",""number_of_canvases"":2,""prompt_style"":""official provider chat template"",""sampler"":""entropy-bounded adaptive"",""sampling"":""replicate count N undisclosed"",""tools"":""none""}",False,tech_report,verified,"Released DiffusionGemma before downstream LoRA. TD-thinking is inherited globally, but the PubMedQA recipe uses a noncanonical 128-token canvas with two canvases."
diffusiongemma-26b-a4b,tau2_bench_airline,69.0,https://arxiv.org/abs/2608.00146,"{""canvas"":256,""context"":""default"",""decoding"":""TD"",""harness"":""official Google cross-model evaluation"",""judge"":""state-based task success"",""max_denoising_steps"":48,""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampler"":""entropy-bounded adaptive"",""sampling"":""trials=4 per task inferred from score totals; N otherwise undisclosed"",""tools"":""benchmark APIs + simulated user""}",False,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
diffusiongemma-26b-a4b,tau2_bench_avg,56.2,https://blog.google/innovation-and-ai/technology/developers-tools/diffusion-gemma-faster-text-generation/,"{""canvas"":256,""context"":""default"",""decoding"":""TD"",""harness"":""official Google launch evaluation"",""judge"":""state-based task success"",""max_denoising_steps"":48,""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampler"":""entropy-bounded adaptive"",""sampling"":""trials=4 per task inferred from score totals; N otherwise undisclosed"",""tools"":""benchmark APIs + simulated user""}",False,official_blog,verified,Google release-blog benchmark chart; highest-priority duplicate for target quality values.
diffusiongemma-26b-a4b,tau2_bench_retail,71.5,https://arxiv.org/abs/2608.00146,"{""canvas"":256,""context"":""default"",""decoding"":""TD"",""harness"":""official Google cross-model evaluation"",""judge"":""state-based task success"",""max_denoising_steps"":48,""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampler"":""entropy-bounded adaptive"",""sampling"":""trials=4 per task inferred from score totals; N otherwise undisclosed"",""tools"":""benchmark APIs + simulated user""}",False,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
diffusiongemma-26b-a4b,tau2_bench_telecom,28.1,https://arxiv.org/abs/2608.00146,"{""canvas"":256,""context"":""default"",""decoding"":""TD"",""harness"":""official Google cross-model evaluation"",""judge"":""state-based task success"",""max_denoising_steps"":48,""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampler"":""entropy-bounded adaptive"",""sampling"":""trials=4 per task inferred from score totals; N otherwise undisclosed"",""tools"":""benchmark APIs + simulated user""}",False,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
gemma-4-26b-a4b,codeelo_rating,1718.0,https://arxiv.org/abs/2608.00146,"{""context"":""default"",""decoding"":""AR MTP"",""harness"":""official Google cross-model evaluation"",""judge"":""Codeforces online judge"",""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampling"":""Google run count undisclosed; canonical samples=8"",""tools"":""none""}",False,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
gemma-4-26b-a4b,gsm8k,96.7,https://arxiv.org/abs/2608.00146,"{""context"":""default"",""decoding"":""AR MTP"",""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampling"":""replicate count N undisclosed"",""tools"":""none""}",True,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
gemma-4-26b-a4b,hiddenmath,87.2,https://arxiv.org/abs/2608.00146,"{""context"":""default"",""decoding"":""AR MTP"",""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampling"":""replicate count N undisclosed"",""tools"":""none""}",True,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
gemma-4-26b-a4b,humaneval,98.8,https://arxiv.org/abs/2608.00146,"{""context"":""default"",""decoding"":""AR MTP"",""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampling"":""replicate count N undisclosed"",""tools"":""none""}",True,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
gemma-4-26b-a4b,lbpp_v2_multilingual,89.5,https://arxiv.org/abs/2608.00146,"{""context"":""default"",""decoding"":""AR MTP"",""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampling"":""replicate count N undisclosed"",""tools"":""none""}",True,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
gemma-4-26b-a4b,mgsm,92.9,https://arxiv.org/abs/2608.00146,"{""context"":""default"",""decoding"":""AR MTP"",""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampling"":""replicate count N undisclosed"",""tools"":""none""}",True,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
gemma-4-26b-a4b,natural2code,96.3,https://arxiv.org/abs/2608.00146,"{""context"":""default"",""decoding"":""AR MTP"",""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampling"":""replicate count N undisclosed"",""tools"":""none""}",True,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
llada-2.1-flash-100b,aime_2026,80.0,https://arxiv.org/abs/2608.00146,"{""block_length"":32,""context"":""default"",""decoding"":""TD S Mode"",""editing_threshold"":0.0,""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""max_post_steps"":16,""mode"":""unknown"",""prompt_style"":""official provider chat template"",""sampling"":""replicate count N undisclosed"",""temperature"":0.0,""thinking_state"":""undisclosed"",""threshold"":0.5,""tools"":""none""}",False,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
llada-2.1-flash-100b,codeelo_rating,718.0,https://arxiv.org/abs/2608.00146,"{""block_length"":32,""context"":""default"",""decoding"":""TD S Mode"",""editing_threshold"":0.0,""harness"":""official Google cross-model evaluation"",""judge"":""Codeforces online judge"",""max_post_steps"":16,""mode"":""unknown"",""prompt_style"":""official provider chat template"",""sampling"":""Google run count undisclosed; canonical samples=8"",""temperature"":0.0,""thinking_state"":""undisclosed"",""threshold"":0.5,""tools"":""none""}",False,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
llada-2.1-flash-100b,gpqa_diamond,68.7,https://arxiv.org/abs/2608.00146,"{""block_length"":32,""context"":""default"",""decoding"":""TD S Mode"",""editing_threshold"":0.0,""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""max_post_steps"":16,""mode"":""unknown"",""prompt_style"":""official provider chat template"",""sampling"":""replicate count N undisclosed"",""temperature"":0.0,""thinking_state"":""undisclosed"",""threshold"":0.5,""tools"":""none""}",False,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
llada-2.1-flash-100b,gsm8k,45.0,https://arxiv.org/abs/2608.00146,"{""block_length"":32,""context"":""default"",""decoding"":""TD S Mode"",""editing_threshold"":0.0,""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""max_post_steps"":16,""mode"":""unknown"",""prompt_style"":""official provider chat template"",""sampling"":""replicate count N undisclosed"",""temperature"":0.0,""thinking_state"":""undisclosed"",""threshold"":0.5,""tools"":""none""}",False,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
llada-2.1-flash-100b,humaneval,90.2,https://arxiv.org/abs/2608.00146,"{""block_length"":32,""context"":""default"",""decoding"":""TD S Mode"",""editing_threshold"":0.0,""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""max_post_steps"":16,""mode"":""unknown"",""prompt_style"":""official provider chat template"",""sampling"":""replicate count N undisclosed"",""temperature"":0.0,""thinking_state"":""undisclosed"",""threshold"":0.5,""tools"":""none""}",False,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
llada-2.1-flash-100b,lbpp_v2_multilingual,45.7,https://arxiv.org/abs/2608.00146,"{""block_length"":32,""context"":""default"",""decoding"":""TD S Mode"",""editing_threshold"":0.0,""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""max_post_steps"":16,""mode"":""unknown"",""prompt_style"":""official provider chat template"",""sampling"":""replicate count N undisclosed"",""temperature"":0.0,""thinking_state"":""undisclosed"",""threshold"":0.5,""tools"":""none""}",False,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
llada-2.1-flash-100b,livecodebench_v6,39.4,https://arxiv.org/abs/2608.00146,"{""block_length"":32,""context"":""default"",""decoding"":""TD S Mode"",""editing_threshold"":0.0,""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""max_post_steps"":16,""mode"":""unknown"",""prompt_style"":""official provider chat template"",""sampling"":""replicate count N undisclosed"",""temperature"":0.0,""thinking_state"":""undisclosed"",""threshold"":0.5,""tools"":""none""}",False,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
llada-2.1-flash-100b,mgsm,6.8,https://arxiv.org/abs/2608.00146,"{""block_length"":32,""context"":""default"",""decoding"":""TD S Mode"",""editing_threshold"":0.0,""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""max_post_steps"":16,""mode"":""unknown"",""prompt_style"":""official provider chat template"",""sampling"":""replicate count N undisclosed"",""temperature"":0.0,""thinking_state"":""undisclosed"",""threshold"":0.5,""tools"":""none""}",False,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
llada-2.1-flash-100b,natural2code,86.9,https://arxiv.org/abs/2608.00146,"{""block_length"":32,""context"":""default"",""decoding"":""TD S Mode"",""editing_threshold"":0.0,""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""max_post_steps"":16,""mode"":""unknown"",""prompt_style"":""official provider chat template"",""sampling"":""replicate count N undisclosed"",""temperature"":0.0,""thinking_state"":""undisclosed"",""threshold"":0.5,""tools"":""none""}",False,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
mercury-2,aime_2026,82.5,https://arxiv.org/abs/2608.00146,"{""context"":""default"",""decoding"":""diffusion"",""effort"":""medium"",""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampling"":""replicate count N undisclosed"",""temperature"":0.75,""tools"":""none""}",True,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
mercury-2,bigbench_extra_hard,43.8,https://arxiv.org/abs/2608.00146,"{""context"":""default"",""decoding"":""diffusion"",""effort"":""medium"",""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampling"":""replicate count N undisclosed"",""temperature"":0.75,""tools"":""none""}",True,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
mercury-2,codeelo_rating,1629.0,https://arxiv.org/abs/2608.00146,"{""context"":""default"",""decoding"":""diffusion"",""effort"":""medium"",""harness"":""official Google cross-model evaluation"",""judge"":""Codeforces online judge"",""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampling"":""Google run count undisclosed; canonical samples=8"",""temperature"":0.75,""tools"":""none""}",False,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
mercury-2,gpqa_diamond,66.7,https://arxiv.org/abs/2608.00146,"{""context"":""default"",""decoding"":""diffusion"",""effort"":""medium"",""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampling"":""replicate count N undisclosed"",""temperature"":0.75,""tools"":""none""}",True,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
mercury-2,gsm8k,95.8,https://arxiv.org/abs/2608.00146,"{""context"":""default"",""decoding"":""diffusion"",""effort"":""medium"",""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampling"":""replicate count N undisclosed"",""temperature"":0.75,""tools"":""none""}",True,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
mercury-2,hiddenmath,82.3,https://arxiv.org/abs/2608.00146,"{""context"":""default"",""decoding"":""diffusion"",""effort"":""medium"",""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampling"":""replicate count N undisclosed"",""temperature"":0.75,""tools"":""none""}",True,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
mercury-2,humaneval,98.2,https://arxiv.org/abs/2608.00146,"{""context"":""default"",""decoding"":""diffusion"",""effort"":""medium"",""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampling"":""replicate count N undisclosed"",""temperature"":0.75,""tools"":""none""}",True,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
mercury-2,ifeval,94.5,https://arxiv.org/abs/2608.00146,"{""context"":""default"",""decoding"":""diffusion"",""effort"":""medium"",""harness"":""official Google cross-model evaluation"",""judge"":""IFEval strict/loose subtype undisclosed"",""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampling"":""replicate count N undisclosed"",""temperature"":0.75,""tools"":""none""}",False,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
mercury-2,lbpp_v2_multilingual,85.0,https://arxiv.org/abs/2608.00146,"{""context"":""default"",""decoding"":""diffusion"",""effort"":""medium"",""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampling"":""replicate count N undisclosed"",""temperature"":0.75,""tools"":""none""}",True,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
mercury-2,livecodebench_v6,74.9,https://arxiv.org/abs/2608.00146,"{""context"":""default"",""decoding"":""diffusion"",""effort"":""medium"",""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampling"":""replicate count N undisclosed"",""temperature"":0.75,""tools"":""none""}",True,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
mercury-2,mgsm,91.2,https://arxiv.org/abs/2608.00146,"{""context"":""default"",""decoding"":""diffusion"",""effort"":""medium"",""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampling"":""replicate count N undisclosed"",""temperature"":0.75,""tools"":""none""}",True,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
mercury-2,mmlu_pro,75.5,https://arxiv.org/abs/2608.00146,"{""context"":""default"",""decoding"":""diffusion"",""effort"":""medium"",""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampling"":""replicate count N undisclosed"",""temperature"":0.75,""tools"":""none""}",True,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
mercury-2,mmmlu,80.6,https://arxiv.org/abs/2608.00146,"{""context"":""default"",""decoding"":""diffusion"",""effort"":""medium"",""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampling"":""replicate count N undisclosed"",""temperature"":0.75,""tools"":""none""}",True,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
mercury-2,natural2code,71.3,https://arxiv.org/abs/2608.00146,"{""context"":""default"",""decoding"":""diffusion"",""effort"":""medium"",""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official provider chat template"",""sampling"":""replicate count N undisclosed"",""temperature"":0.75,""tools"":""none""}",True,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
nemotron-diffusion-14b,aime_2026,40.0,https://arxiv.org/abs/2608.00146,"{""block_length"":32,""context"":""default"",""decoding"":""TD Diffusion Mode"",""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""mode"":""unknown"",""prompt_style"":""official provider chat template"",""sampling"":""replicate count N undisclosed"",""thinking_state"":""undisclosed"",""threshold"":0.9,""tools"":""none""}",False,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
nemotron-diffusion-14b,gpqa_diamond,47.0,https://arxiv.org/abs/2608.00146,"{""block_length"":32,""context"":""default"",""decoding"":""TD Diffusion Mode"",""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""mode"":""unknown"",""prompt_style"":""official provider chat template"",""sampling"":""replicate count N undisclosed"",""thinking_state"":""undisclosed"",""threshold"":0.9,""tools"":""none""}",False,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
nemotron-diffusion-14b,hiddenmath,44.3,https://arxiv.org/abs/2608.00146,"{""block_length"":32,""context"":""default"",""decoding"":""TD Diffusion Mode"",""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""mode"":""unknown"",""prompt_style"":""official provider chat template"",""sampling"":""replicate count N undisclosed"",""thinking_state"":""undisclosed"",""threshold"":0.9,""tools"":""none""}",False,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
nemotron-diffusion-14b,humaneval,86.0,https://arxiv.org/abs/2608.00146,"{""block_length"":32,""context"":""default"",""decoding"":""TD Diffusion Mode"",""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""mode"":""unknown"",""prompt_style"":""official provider chat template"",""sampling"":""replicate count N undisclosed"",""thinking_state"":""undisclosed"",""threshold"":0.9,""tools"":""none""}",False,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
nemotron-diffusion-14b,ifeval,72.1,https://arxiv.org/abs/2608.00146,"{""block_length"":32,""context"":""default"",""decoding"":""TD Diffusion Mode"",""harness"":""official Google cross-model evaluation"",""judge"":""IFEval strict/loose subtype undisclosed"",""mode"":""unknown"",""prompt_style"":""official provider chat template"",""sampling"":""replicate count N undisclosed"",""thinking_state"":""undisclosed"",""threshold"":0.9,""tools"":""none""}",False,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
nemotron-diffusion-14b,lbpp_v2_multilingual,40.7,https://arxiv.org/abs/2608.00146,"{""block_length"":32,""context"":""default"",""decoding"":""TD Diffusion Mode"",""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""mode"":""unknown"",""prompt_style"":""official provider chat template"",""sampling"":""replicate count N undisclosed"",""thinking_state"":""undisclosed"",""threshold"":0.9,""tools"":""none""}",False,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
nemotron-diffusion-14b,livecodebench_v6,28.6,https://arxiv.org/abs/2608.00146,"{""block_length"":32,""context"":""default"",""decoding"":""TD Diffusion Mode"",""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""mode"":""unknown"",""prompt_style"":""official provider chat template"",""sampling"":""replicate count N undisclosed"",""thinking_state"":""undisclosed"",""threshold"":0.9,""tools"":""none""}",False,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
nemotron-diffusion-14b,mgsm,69.3,https://arxiv.org/abs/2608.00146,"{""block_length"":32,""context"":""default"",""decoding"":""TD Diffusion Mode"",""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""mode"":""unknown"",""prompt_style"":""official provider chat template"",""sampling"":""replicate count N undisclosed"",""thinking_state"":""undisclosed"",""threshold"":0.9,""tools"":""none""}",False,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
nemotron-diffusion-14b,natural2code,73.3,https://arxiv.org/abs/2608.00146,"{""block_length"":32,""context"":""default"",""decoding"":""TD Diffusion Mode"",""harness"":""official Google cross-model evaluation"",""judge"":""benchmark-specified"",""mode"":""unknown"",""prompt_style"":""official provider chat template"",""sampling"":""replicate count N undisclosed"",""thinking_state"":""undisclosed"",""threshold"":0.9,""tools"":""none""}",False,tech_report,verified,DiffusionGemma Technical Report Table 3; replicate count N is undisclosed.
claude-fable-5,apex_swe,58.8,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max, with fallback"",""harness"":""Mercor APEX-SWE harness"",""judge"":""Mercor integration and observability task verifier"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic software-engineering environment""}",True,model_card,verified,"card46 APEX-SWE Pass@1 (%); source effort=max, with fallback; benchmark protocol matches canonical=True. Unknown source fields remain unknown."
claude-fable-5,deep_swe_v1_0,66.1,https://x.ai/news/grok-4-5,"{""context"":""unknown"",""effort"":""max"",""harness"":""provider harnesses run by Artificial Analysis"",""judge"":""behavioral and correctness verifiers"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic repository shell/editor""}",False,official_blog,verified,release45 DeepSWE 1.0 Pass@1 (%); source effort=max; benchmark protocol matches canonical=False. Unknown source fields remain unknown.
claude-fable-5,eebench_v1_core_reward,54.2,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max, with fallback"",""harness"":""EEBench v1 core public leaderboard harness"",""judge"":""physical correctness and functionality reward"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""electrical-engineering design tools""}",True,model_card,verified,"card46 EEBench v1 core Reward (%); source effort=max, with fallback; benchmark protocol matches canonical=True. Unknown source fields remain unknown."
claude-fable-5,gdpval_aa_v2_elo,1741.0,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max, with fallback"",""harness"":""Artificial Analysis GDPval-AA v2 harness"",""judge"":""pairwise Bradley-Terry Elo"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""benchmark-specified knowledge-work tools""}",True,model_card,verified,"card46 AA GDPVal v2 AA GDPVal Elo; source effort=max, with fallback; benchmark protocol matches canonical=True. Unknown source fields remain unknown."
claude-fable-5,harvey_lab_vals_final_score,14.2,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max, with fallback"",""harness"":""Vals implementation of Harvey LAB"",""judge"":""expert all-pass final score"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""Vals legal-agent file tools""}",True,model_card,verified,"card46 Harvey Legal Agent Benchmark Final score (%); source effort=max, with fallback; benchmark protocol matches canonical=True. Unknown source fields remain unknown."
claude-fable-5,parametric_cad_bench_combined,88.5,https://media.x.ai/v1/website/4p5-5184fdf9.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""gNucleus Parametric CAD Bench harness"",""judge"":""combined executability and geometric reward"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""parametric CAD agent tools""}",False,model_card,verified,card45 Parametric CAD Bench Reward (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
claude-fable-5,swe_atlas_codebase_qna,83.0,https://media.x.ai/v1/website/4p5-5184fdf9.pdf,"{""context"":""unknown"",""effort"":""max, with fallback"",""harness"":""SWE-Atlas-QnA tracked split"",""judge"":""repository-ground-truth rubric"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""codebase reading and exploration""}",True,model_card,verified,"card45 SWE-Atlas-QnA Accuracy (%); source effort=max, with fallback; benchmark protocol matches canonical=True. Unknown source fields remain unknown."
claude-fable-5,swe_marathon_v1_0,24.0,https://media.x.ai/v1/website/4p5-5184fdf9.pdf,"{""context"":""unknown"",""effort"":""max, with fallback"",""harness"":""SWE-Marathon v1.0 harness"",""judge"":""multi-layer resolution verification"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic code execution""}",True,model_card,verified,"card45 SWE-Marathon v1.0 Resolution rate (%); source effort=max, with fallback; benchmark protocol matches canonical=True. Unknown source fields remain unknown."
claude-fable-5,swe_marathon_v1_1,45.0,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max, with fallback"",""harness"":""SWE-Marathon v1.1 harness"",""judge"":""multi-layer resolution verification"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic code execution""}",True,model_card,verified,"card46 SWE-Marathon v1.1 Resolution rate (%); source effort=max, with fallback; benchmark protocol matches canonical=True. Unknown source fields remain unknown."
claude-fable-5,terminal_bench_3_0,34.1,https://x.ai/news/grok-4-6,"{""context"":""unknown"",""effort"":""max"",""harness"":""Terminal-Bench 3.0 / FrontierBench fixed harness"",""judge"":""verified task-success evaluator"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""container terminal""}",False,official_blog,verified,release46 Terminal-Bench v3.0 Task success rate (%); source effort=max; benchmark protocol matches canonical=False. Unknown source fields remain unknown.
claude-opus-4.7,deep_swe_v1_0,40.12,https://x.ai/news/grok-4-5,"{""context"":""unknown"",""effort"":""max"",""harness"":""provider harnesses run by Artificial Analysis"",""judge"":""behavioral and correctness verifiers"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic repository shell/editor""}",False,official_blog,verified,release45 DeepSWE 1.0 Pass@1 (%); source effort=max; benchmark protocol matches canonical=False. Unknown source fields remain unknown.
claude-opus-4.8,apex_swe,47.3,https://media.x.ai/v1/website/4p5-5184fdf9.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""Mercor APEX-SWE harness"",""judge"":""Mercor integration and observability task verifier"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic software-engineering environment""}",True,model_card,verified,card45 APEX-SWE Pass@1 (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
claude-opus-4.8,deep_swe_v1_0,55.75,https://x.ai/news/grok-4-5,"{""context"":""unknown"",""effort"":""max"",""harness"":""provider harnesses run by Artificial Analysis"",""judge"":""behavioral and correctness verifiers"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic repository shell/editor""}",False,official_blog,verified,release45 DeepSWE 1.0 Pass@1 (%); source effort=max; benchmark protocol matches canonical=False. Unknown source fields remain unknown.
claude-opus-4.8,deepsearchqa_acc,84.8,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""SpaceXAI internal DeepSearchQA implementation"",""judge"":""DeepSearchQA answer-accuracy grader"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""web search agent""}",False,model_card,verified,card46 DeepSearchQA Accuracy (%); source effort=max; benchmark protocol matches canonical=False. Unknown source fields remain unknown.
claude-opus-4.8,eebench_v1_core_reward,51.4,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""EEBench v1 core public leaderboard harness"",""judge"":""physical correctness and functionality reward"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""electrical-engineering design tools""}",False,model_card,verified,card46 EEBench v1 core Reward (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
claude-opus-4.8,falseclaimbench_accuracy,59.0,https://media.x.ai/v1/website/4p5-5184fdf9.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""SpaceXAI internal FalseClaimBench harness"",""judge"":""claim-to-final-workspace-state verifier"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""workspace editing and command tools""}",False,model_card,verified,card45 FalseClaimBench Accuracy (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
claude-opus-4.8,hackerbench_v0_2_benign_refusal,0.0,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""HackerBench v0.2 with standard safeguards"",""judge"":""benign refusal grader"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""constrained cyber task tools""}",False,model_card,verified,card46 HackerBench v0.2 Benign Refusal (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
claude-opus-4.8,hackerbench_v0_2_harmful_compliance,22.8,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""HackerBench v0.2 with standard safeguards"",""judge"":""harmful or dual-use compliance grader"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""constrained cyber task tools""}",False,model_card,verified,card46 HackerBench v0.2 Harmful/Dual-Use Compliance (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
claude-opus-4.8,parametric_cad_bench_combined,88.2,https://media.x.ai/v1/website/4p5-5184fdf9.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""gNucleus Parametric CAD Bench harness"",""judge"":""combined executability and geometric reward"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""parametric CAD agent tools""}",False,model_card,verified,card45 Parametric CAD Bench Reward (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
claude-opus-4.8,single_turn_hallucination_rate,3.4,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""SpaceXAI internal single-turn factuality harness"",""judge"":""separate factual-support grader"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""none""}",False,model_card,verified,card46 Single-turn hallucination Hallucination rate (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
claude-opus-4.8,spacexai_mts_accuracy,55.8,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""Grok Build"",""judge"":""task correctness under fixed rollout budget"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""Grok Build coding and research tools""}",False,model_card,verified,card46 SpaceXAI MTS Eval Correct (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
claude-opus-4.8,swe_atlas_codebase_qna,82.0,https://media.x.ai/v1/website/4p5-5184fdf9.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""SWE-Atlas-QnA tracked split"",""judge"":""repository-ground-truth rubric"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""codebase reading and exploration""}",False,model_card,verified,card45 SWE-Atlas-QnA Accuracy (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
claude-opus-4.8,swe_marathon_v1_1,48.8,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""SWE-Marathon v1.1 harness"",""judge"":""multi-layer resolution verification"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic code execution""}",False,model_card,verified,card46 SWE-Marathon v1.1 Resolution rate (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
claude-opus-4.8,terminal_bench_3_0,21.1,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""Terminal-Bench 3.0 / FrontierBench fixed harness"",""judge"":""verified task-success evaluator"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""container terminal""}",False,model_card,verified,card46 Terminal-Bench 3.0 Task success rate (%); source effort=max; benchmark protocol matches canonical=False. Unknown source fields remain unknown.
claude-opus-4.8,xai_inferenceeval_accuracy,40.7,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""SpaceXAI InferenceEval harness"",""judge"":""hidden GPU unit score gated by integration probe"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""Grok Build coding tools and GPU tests""}",False,model_card,verified,card46 InferenceEval Accuracy (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
claude-opus-4.8,xai_kernelbenchinternal_accuracy_efficiency,67.6,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""SpaceXAI KernelBenchInternal harness"",""judge"":""correctness and measured speedup composite"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""sandboxed GPU coding tools""}",False,model_card,verified,card46 KernelBenchInternal Accuracy / efficiency (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
claude-opus-5,apex_agents,60.6,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""Mercor APEX-Agents harness"",""judge"":""expert all-criteria-pass rubric"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""multi-application professional tools""}",False,model_card,verified,card46 APEX-Agents Pass@1 (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
claude-opus-5,apex_swe,63.7,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""Mercor APEX-SWE harness"",""judge"":""Mercor integration and observability task verifier"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic software-engineering environment""}",False,model_card,verified,card46 APEX-SWE Pass@1 (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
claude-opus-5,cadgenbench_generation_cad_score,36.6,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""CadGenBench generation harness"",""judge"":""generation CAD score"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""CAD generation agent tools""}",False,model_card,verified,card46 CadGenBench Generation Reward (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
claude-opus-5,eebench_v1_core_reward,61.6,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""EEBench v1 core public leaderboard harness"",""judge"":""physical correctness and functionality reward"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""electrical-engineering design tools""}",False,model_card,verified,card46 EEBench v1 core Reward (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
claude-opus-5,gdpval_aa_v2_elo,1849.0,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""Artificial Analysis GDPval-AA v2 harness"",""judge"":""pairwise Bradley-Terry Elo"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""benchmark-specified knowledge-work tools""}",False,model_card,verified,card46 AA GDPVal v2 AA GDPVal Elo; source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
claude-opus-5,harvey_lab_vals_final_score,11.7,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""Vals implementation of Harvey LAB"",""judge"":""expert all-pass final score"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""Vals legal-agent file tools""}",False,model_card,verified,card46 Harvey Legal Agent Benchmark Final score (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
claude-opus-5,parametric_cad_bench_combined,90.6,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""gNucleus Parametric CAD Bench harness"",""judge"":""combined executability and geometric reward"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""parametric CAD agent tools""}",False,model_card,verified,card46 Parametric CAD Bench Reward (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
claude-opus-5,spacexai_mts_accuracy,52.6,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""Grok Build"",""judge"":""task correctness under fixed rollout budget"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""Grok Build coding and research tools""}",False,model_card,verified,card46 SpaceXAI MTS Eval Correct (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
claude-opus-5,swe_marathon_v1_1,50.0,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""SWE-Marathon v1.1 harness"",""judge"":""multi-layer resolution verification"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic code execution""}",False,model_card,verified,card46 SWE-Marathon v1.1 Resolution rate (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
claude-opus-5,terminal_bench_3_0,43.5,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""Terminal-Bench 3.0 / FrontierBench fixed harness"",""judge"":""verified task-success evaluator"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""container terminal""}",False,model_card,verified,card46 Terminal-Bench 3.0 Task success rate (%); source effort=max; benchmark protocol matches canonical=False. Unknown source fields remain unknown.
claude-opus-5,vals_index,74.8,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""Vals Index live harness"",""judge"":""Vals weighted suite score"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""benchmark-specified professional tools""}",False,model_card,verified,card46 Vals Index Suite score (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
claude-opus-5,xai_inferenceeval_accuracy,43.0,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""SpaceXAI InferenceEval harness"",""judge"":""hidden GPU unit score gated by integration probe"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""Grok Build coding tools and GPU tests""}",False,model_card,verified,card46 InferenceEval Accuracy (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
claude-opus-5,xai_kernelbenchinternal_accuracy_efficiency,73.0,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""SpaceXAI KernelBenchInternal harness"",""judge"":""correctness and measured speedup composite"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""sandboxed GPU coding tools""}",False,model_card,verified,card46 KernelBenchInternal Accuracy / efficiency (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
claude-sonnet-5,apex_agents,32.5,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""Mercor APEX-Agents harness"",""judge"":""expert all-criteria-pass rubric"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""multi-application professional tools""}",True,model_card,verified,card46 APEX-Agents Pass@1 (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
claude-sonnet-5,apex_swe,43.6,https://media.x.ai/v1/website/4p5-5184fdf9.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""Mercor APEX-SWE harness"",""judge"":""Mercor integration and observability task verifier"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic software-engineering environment""}",True,model_card,verified,card45 APEX-SWE Pass@1 (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
claude-sonnet-5,frontiercode_extended_v1_1,56.2,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""FrontierCode v1.1 Extended harness"",""judge"":""maintainer-authored rubrics, verifiers, and hard blockers"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""fixed coding-agent scaffold""}",False,model_card,verified,card46 FrontierCode v1.1 Extended Extended score (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
claude-sonnet-5,parametric_cad_bench_combined,86.9,https://media.x.ai/v1/website/4p5-5184fdf9.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""gNucleus Parametric CAD Bench harness"",""judge"":""combined executability and geometric reward"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""parametric CAD agent tools""}",False,model_card,verified,card45 Parametric CAD Bench Reward (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
claude-sonnet-5,swe_marathon_v1_1,30.0,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""SWE-Marathon v1.1 harness"",""judge"":""multi-layer resolution verification"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic code execution""}",False,model_card,verified,card46 SWE-Marathon v1.1 Resolution rate (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
claude-sonnet-5,tau3_banking,28.2,https://media.x.ai/v1/website/4p5-5184fdf9.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""Artificial Analysis tau3-banking harness"",""judge"":""final-state task accuracy"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""banking APIs and simulated user""}",False,model_card,verified,card45 tau3-banking Accuracy (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
claude-sonnet-5,terminal_bench_3_0,14.6,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""Terminal-Bench 3.0 / FrontierBench fixed harness"",""judge"":""verified task-success evaluator"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""container terminal""}",False,model_card,verified,card46 Terminal-Bench 3.0 Task success rate (%); source effort=max; benchmark protocol matches canonical=False. Unknown source fields remain unknown.
claude-sonnet-5,vals_index,68.6,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""Vals Index live harness"",""judge"":""Vals weighted suite score"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""benchmark-specified professional tools""}",False,model_card,verified,card46 Vals Index Suite score (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
glm-5.2,apex_swe,37.3,https://media.x.ai/v1/website/4p5-5184fdf9.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""Mercor APEX-SWE harness"",""judge"":""Mercor integration and observability task verifier"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic software-engineering environment""}",True,model_card,verified,card45 APEX-SWE Pass@1 (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
glm-5.2,falseclaimbench_accuracy,38.0,https://media.x.ai/v1/website/4p5-5184fdf9.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""SpaceXAI internal FalseClaimBench harness"",""judge"":""claim-to-final-workspace-state verifier"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""workspace editing and command tools""}",True,model_card,verified,card45 FalseClaimBench Accuracy (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
glm-5.2,gdpval_aa_v2_elo,1514.0,https://media.x.ai/v1/website/4p5-5184fdf9.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""Artificial Analysis GDPval-AA v2 harness"",""judge"":""pairwise Bradley-Terry Elo"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""benchmark-specified knowledge-work tools""}",True,model_card,verified,card45 GDPval-AA v2 AA score; source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
glm-5.2,swe_atlas_codebase_qna,74.0,https://media.x.ai/v1/website/4p5-5184fdf9.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""SWE-Atlas-QnA tracked split"",""judge"":""repository-ground-truth rubric"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""codebase reading and exploration""}",True,model_card,verified,card45 SWE-Atlas-QnA Accuracy (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
gpt-5.5,apex_swe,37.0,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""xhigh"",""harness"":""Mercor APEX-SWE harness"",""judge"":""Mercor integration and observability task verifier"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic software-engineering environment""}",True,model_card,verified,card46 APEX-SWE Pass@1 (%); source effort=xhigh; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
gpt-5.5,deep_swe_v1_0,64.31,https://x.ai/news/grok-4-5,"{""context"":""unknown"",""effort"":""xhigh"",""harness"":""provider harnesses run by Artificial Analysis"",""judge"":""behavioral and correctness verifiers"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic repository shell/editor""}",False,official_blog,verified,release45 DeepSWE 1.0 Pass@1 (%); source effort=xhigh; benchmark protocol matches canonical=False. Unknown source fields remain unknown.
gpt-5.5,deepsearchqa_acc,63.7,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""xhigh"",""harness"":""SpaceXAI internal DeepSearchQA implementation"",""judge"":""DeepSearchQA answer-accuracy grader"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""web search agent""}",False,model_card,verified,card46 DeepSearchQA Accuracy (%); source effort=xhigh; benchmark protocol matches canonical=False. Unknown source fields remain unknown.
gpt-5.5,eebench_v1_core_reward,42.3,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""xhigh"",""harness"":""EEBench v1 core public leaderboard harness"",""judge"":""physical correctness and functionality reward"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""electrical-engineering design tools""}",True,model_card,verified,card46 EEBench v1 core Reward (%); source effort=xhigh; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
gpt-5.5,frontiercode_extended_v1_1,56.7,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""xhigh"",""harness"":""FrontierCode v1.1 Extended harness"",""judge"":""maintainer-authored rubrics, verifiers, and hard blockers"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""fixed coding-agent scaffold""}",True,model_card,verified,card46 FrontierCode v1.1 Extended Extended score (%); source effort=xhigh; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
gpt-5.5,hackerbench_v0_2_benign_refusal,1.1,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""xhigh"",""harness"":""HackerBench v0.2 with standard safeguards"",""judge"":""benign refusal grader"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""constrained cyber task tools""}",True,model_card,verified,card46 HackerBench v0.2 Benign Refusal (%); source effort=xhigh; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
gpt-5.5,hackerbench_v0_2_harmful_compliance,25.0,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""xhigh"",""harness"":""HackerBench v0.2 with standard safeguards"",""judge"":""harmful or dual-use compliance grader"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""constrained cyber task tools""}",True,model_card,verified,card46 HackerBench v0.2 Harmful/Dual-Use Compliance (%); source effort=xhigh; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
gpt-5.5,single_turn_hallucination_rate,1.1,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""xhigh"",""harness"":""SpaceXAI internal single-turn factuality harness"",""judge"":""separate factual-support grader"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""none""}",True,model_card,verified,card46 Single-turn hallucination Hallucination rate (%); source effort=xhigh; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
gpt-5.5,spacexai_mts_accuracy,46.4,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""xhigh"",""harness"":""Grok Build"",""judge"":""task correctness under fixed rollout budget"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""Grok Build coding and research tools""}",True,model_card,verified,card46 SpaceXAI MTS Eval Correct (%); source effort=xhigh; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
gpt-5.5,swe_atlas_codebase_qna,81.0,https://media.x.ai/v1/website/4p5-5184fdf9.pdf,"{""context"":""unknown"",""effort"":""xhigh"",""harness"":""SWE-Atlas-QnA tracked split"",""judge"":""repository-ground-truth rubric"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""codebase reading and exploration""}",True,model_card,verified,card45 SWE-Atlas-QnA Accuracy (%); source effort=xhigh; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
gpt-5.5,xai_inferenceeval_accuracy,38.9,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""xhigh"",""harness"":""SpaceXAI InferenceEval harness"",""judge"":""hidden GPU unit score gated by integration probe"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""Grok Build coding tools and GPU tests""}",True,model_card,verified,card46 InferenceEval Accuracy (%); source effort=xhigh; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
gpt-5.5,xai_securecodereview_reward,64.1,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""xhigh"",""harness"":""SpaceXAI SecureCodeReview harness"",""judge"":""security-fix reward with regression checks"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""fixed secure-code-review tools""}",True,model_card,verified,card46 SecureCodeReview Reward (%); source effort=xhigh; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
gpt-5.6-sol,apex_swe,45.8,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""xhigh"",""harness"":""Mercor APEX-SWE harness"",""judge"":""Mercor integration and observability task verifier"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic software-engineering environment""}",False,model_card,verified,card46 APEX-SWE Pass@1 (%); source effort=xhigh; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
gpt-5.6-sol,cadgenbench_generation_cad_score,37.1,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""xhigh"",""harness"":""CadGenBench generation harness"",""judge"":""generation CAD score"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""CAD generation agent tools""}",False,model_card,verified,card46 CadGenBench Generation Reward (%); source effort=xhigh; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
gpt-5.6-sol,cybergym,83.6,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""CyberGym unrestricted capability harness"",""judge"":""PoC reproduced on vulnerable but not fixed version"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unrestricted agentic code execution""}",True,model_card,verified,card46 CyberGym Mean Reproduced (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
gpt-5.6-sol,deepsearchqa_acc,75.0,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""SpaceXAI internal DeepSearchQA implementation"",""judge"":""DeepSearchQA answer-accuracy grader"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""web search agent""}",False,model_card,verified,card46 DeepSearchQA Accuracy (%); source effort=max; benchmark protocol matches canonical=False. Unknown source fields remain unknown.
gpt-5.6-sol,eebench_v1_core_reward,39.4,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""EEBench v1 core public leaderboard harness"",""judge"":""physical correctness and functionality reward"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""electrical-engineering design tools""}",True,model_card,verified,card46 EEBench v1 core Reward (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
gpt-5.6-sol,gdpval_aa_v2_elo,1728.0,https://x.ai/news/grok-4-6,"{""context"":""unknown"",""effort"":""max"",""harness"":""Artificial Analysis GDPval-AA v2 harness"",""judge"":""pairwise Bradley-Terry Elo"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""benchmark-specified knowledge-work tools""}",True,official_blog,verified,release46 GDPVal-AA v2 Elo; source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
gpt-5.6-sol,hackerbench_v0_2_benign_refusal,0.0,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""HackerBench v0.2 with standard safeguards"",""judge"":""benign refusal grader"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""constrained cyber task tools""}",True,model_card,verified,card46 HackerBench v0.2 Benign Refusal (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
gpt-5.6-sol,hackerbench_v0_2_harmful_compliance,35.7,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""HackerBench v0.2 with standard safeguards"",""judge"":""harmful or dual-use compliance grader"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""constrained cyber task tools""}",True,model_card,verified,card46 HackerBench v0.2 Harmful/Dual-Use Compliance (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
gpt-5.6-sol,harvey_lab_vals_final_score,2.5,https://x.ai/news/grok-4-6,"{""context"":""unknown"",""effort"":""max"",""harness"":""Vals implementation of Harvey LAB"",""judge"":""expert all-pass final score"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""Vals legal-agent file tools""}",True,official_blog,verified,release46 Harvey LAB (Vals) Final score (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
gpt-5.6-sol,parametric_cad_bench_combined,86.5,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""gNucleus Parametric CAD Bench harness"",""judge"":""combined executability and geometric reward"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""parametric CAD agent tools""}",True,model_card,verified,card46 Parametric CAD Bench Reward (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
gpt-5.6-sol,spacexai_mts_accuracy,52.1,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""xhigh"",""harness"":""Grok Build"",""judge"":""task correctness under fixed rollout budget"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""Grok Build coding and research tools""}",False,model_card,verified,card46 SpaceXAI MTS Eval Correct (%); source effort=xhigh; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
gpt-5.6-sol,swe_atlas_codebase_qna,84.0,https://media.x.ai/v1/website/4p5-5184fdf9.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""SWE-Atlas-QnA tracked split"",""judge"":""repository-ground-truth rubric"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""codebase reading and exploration""}",True,model_card,verified,card45 SWE-Atlas-QnA Accuracy (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
gpt-5.6-sol,swe_marathon_v1_1,42.5,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""SWE-Marathon v1.1 harness"",""judge"":""multi-layer resolution verification"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic code execution""}",True,model_card,verified,card46 SWE-Marathon v1.1 Resolution rate (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
gpt-5.6-sol,terminal_bench_3_0,34.6,https://x.ai/news/grok-4-6,"{""context"":""unknown"",""effort"":""max"",""harness"":""Terminal-Bench 3.0 / FrontierBench fixed harness"",""judge"":""verified task-success evaluator"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""container terminal""}",False,official_blog,verified,release46 Terminal-Bench v3.0 Task success rate (%); source effort=max; benchmark protocol matches canonical=False. Unknown source fields remain unknown.
gpt-5.6-sol,xai_inferenceeval_accuracy,44.1,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""SpaceXAI InferenceEval harness"",""judge"":""hidden GPU unit score gated by integration probe"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""Grok Build coding tools and GPU tests""}",True,model_card,verified,card46 InferenceEval Accuracy (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
gpt-5.6-sol,xai_kernelbenchinternal_accuracy_efficiency,73.1,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""SpaceXAI KernelBenchInternal harness"",""judge"":""correctness and measured speedup composite"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""sandboxed GPU coding tools""}",True,model_card,verified,card46 KernelBenchInternal Accuracy / efficiency (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
gpt-5.6-sol,xai_securecodereview_reward,57.5,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""SpaceXAI SecureCodeReview harness"",""judge"":""security-fix reward with regression checks"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""fixed secure-code-review tools""}",True,model_card,verified,card46 SecureCodeReview Reward (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
gpt-5.6-terra,gdpval_aa_v2_elo,1578.0,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""Artificial Analysis GDPval-AA v2 harness"",""judge"":""pairwise Bradley-Terry Elo"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""benchmark-specified knowledge-work tools""}",True,model_card,verified,card46 AA GDPVal v2 AA GDPVal Elo; source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
gpt-5.6-terra,swe_atlas_codebase_qna,81.0,https://media.x.ai/v1/website/4p5-5184fdf9.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""SWE-Atlas-QnA tracked split"",""judge"":""repository-ground-truth rubric"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""codebase reading and exploration""}",True,model_card,verified,card45 SWE-Atlas-QnA Accuracy (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
gpt-5.6-terra,tau3_banking,31.8,https://media.x.ai/v1/website/4p5-5184fdf9.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""Artificial Analysis tau3-banking harness"",""judge"":""final-state task accuracy"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""banking APIs and simulated user""}",True,model_card,verified,card45 tau3-banking Accuracy (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
gpt-5.6-terra,vals_index,65.1,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""Vals Index live harness"",""judge"":""Vals weighted suite score"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""benchmark-specified professional tools""}",True,model_card,verified,card46 Vals Index Suite score (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.3,gdpval_aa_v2_elo,1085.0,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""unknown"",""harness"":""Artificial Analysis GDPval-AA v2 harness"",""judge"":""pairwise Bradley-Terry Elo"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""benchmark-specified knowledge-work tools""}",False,model_card,verified,card46 AA GDPVal v2 AA GDPVal Elo; source effort=unknown; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.3,spacexai_mts_accuracy,40.6,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""unknown"",""harness"":""Grok Build"",""judge"":""task correctness under fixed rollout budget"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""Grok Build coding and research tools""}",False,model_card,verified,card46 SpaceXAI MTS Eval Correct (%); source effort=unknown; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,aa_briefcase_elo,1313.0,https://x.ai/news/grok-4-6,"{""context"":""unknown"",""effort"":""high"",""harness"":""Artificial Analysis AA-Briefcase harness"",""judge"":""rubric plus analytical and presentation pairwise Elo"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""offline multi-file knowledge-work tools""}",True,official_blog,verified,release46 AA-Briefcase Elo; source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,aa_intelligence_index_v4_1,56.0,https://x.ai/news/grok-4-6,"{""context"":""unknown"",""effort"":""high"",""harness"":""Artificial Analysis live v4.1 harness"",""judge"":""Artificial Analysis weighted composite"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""benchmark-specified""}",True,official_blog,verified,release46 AA Intelligence Index Composite score; source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,apex_agents,47.1,https://x.ai/news/grok-4-6,"{""context"":""unknown"",""effort"":""high"",""harness"":""Mercor APEX-Agents harness"",""judge"":""expert all-criteria-pass rubric"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""multi-application professional tools""}",True,official_blog,verified,release46 APEX-Agents Pass@1 (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,apex_swe,53.6,https://x.ai/news/grok-4-6,"{""context"":""unknown"",""effort"":""high"",""harness"":""Mercor APEX-SWE harness"",""judge"":""Mercor integration and observability task verifier"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic software-engineering environment""}",True,official_blog,verified,release46 APEX-SWE Pass@1 (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,bixbench,93.8,https://media.x.ai/v1/website/4p5-5184fdf9.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""Grok Build default tool-on BixBench harness"",""judge"":""zero-shot multiple-choice accuracy"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""analysis tools enabled by default""}",True,model_card,verified,card45 BixBench zero-shot MCQ Accuracy (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,cadgenbench_generation_cad_score,33.2,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""CadGenBench generation harness"",""judge"":""generation CAD score"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""CAD generation agent tools""}",True,model_card,verified,card46 CadGenBench Generation Reward (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,cursorbench_3_2,66.7,https://x.ai/news/grok-4-6,"{""context"":""unknown"",""effort"":""high"",""harness"":""Cursor first-party agent harness"",""judge"":""CursorBench task score"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""Cursor coding-agent tools""}",True,official_blog,verified,release46 CursorBench v3.2 Score (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,cve_bench_reward,35.2,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""CVE-Bench sandbox"",""judge"":""CVE-Bench exploit reward"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unrestricted sandboxed web-application tools""}",True,model_card,verified,card46 CVE-Bench Reward (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,cybergym,79.0,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""CyberGym unrestricted capability harness"",""judge"":""PoC reproduced on vulnerable but not fixed version"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unrestricted agentic code execution""}",True,model_card,verified,card46 CyberGym Mean Reproduced (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,deep_swe_v1_0,62.0,https://x.ai/news/grok-4-5,"{""context"":""unknown"",""effort"":""unknown"",""harness"":""provider harnesses run by Artificial Analysis"",""judge"":""behavioral and correctness verifiers"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic repository shell/editor""}",False,official_blog,verified,release45 DeepSWE 1.0 Pass@1 (%); source effort=unknown; benchmark protocol matches canonical=False. Unknown source fields remain unknown.
grok-4.5,deepsearchqa_acc,38.4,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""SpaceXAI internal DeepSearchQA implementation"",""judge"":""DeepSearchQA answer-accuracy grader"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""web search agent""}",False,model_card,verified,card46 DeepSearchQA Accuracy (%); source effort=high; benchmark protocol matches canonical=False. Unknown source fields remain unknown.
grok-4.5,eebench_v1_core_reward,50.9,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""EEBench v1 core public leaderboard harness"",""judge"":""physical correctness and functionality reward"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""electrical-engineering design tools""}",True,model_card,verified,card46 EEBench v1 core Reward (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,falseclaimbench_accuracy,84.0,https://media.x.ai/v1/website/4p5-5184fdf9.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""SpaceXAI internal FalseClaimBench harness"",""judge"":""claim-to-final-workspace-state verifier"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""workspace editing and command tools""}",True,model_card,verified,card45 FalseClaimBench Accuracy (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,frontier_swe,78.0,https://media.x.ai/v1/website/4p5-5184fdf9.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""public FrontierSWE harness; Grok CLI for Grok 4.5"",""judge"":""continuous partial-credit task scoring and dominance"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""mean@5"",""temperature"":""unknown"",""tools"":""agentic code execution""}",True,model_card,verified,card45 FrontierSWE Dominance (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,frontiercode_extended_v1_1,56.6,https://x.ai/news/grok-4-6,"{""context"":""unknown"",""effort"":""high"",""harness"":""FrontierCode v1.1 Extended harness"",""judge"":""maintainer-authored rubrics, verifiers, and hard blockers"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""fixed coding-agent scaffold""}",True,official_blog,verified,release46 FrontierCode v1.1 Extended Extended score (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,hackerbench_v0_2_benign_refusal,0.0,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""HackerBench v0.2 with standard safeguards"",""judge"":""benign refusal grader"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""constrained cyber task tools""}",True,model_card,verified,card46 HackerBench v0.2 Benign Refusal (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,hackerbench_v0_2_harmful_compliance,7.8,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""HackerBench v0.2 with standard safeguards"",""judge"":""harmful or dual-use compliance grader"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""constrained cyber task tools""}",True,model_card,verified,card46 HackerBench v0.2 Harmful/Dual-Use Compliance (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,harvey_lab_vals_final_score,12.9,https://x.ai/news/grok-4-6,"{""context"":""unknown"",""effort"":""high"",""harness"":""Vals implementation of Harvey LAB"",""judge"":""expert all-pass final score"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""Vals legal-agent file tools""}",True,official_blog,verified,release46 Harvey LAB (Vals) Final score (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,lab_bench_protocolqa_accuracy,87.0,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""LAB-Bench ProtocolQA harness"",""judge"":""open-ended single-error accuracy grader"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""none; no safeguards""}",False,model_card,verified,card46 ProtocolQA Open-Ended Accuracy (%); source effort=high; benchmark protocol matches canonical=False. Unknown source fields remain unknown.
grok-4.5,parametric_cad_bench_combined,83.7,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""gNucleus Parametric CAD Bench harness"",""judge"":""combined executability and geometric reward"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""parametric CAD agent tools""}",True,model_card,verified,card46 Parametric CAD Bench Reward (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,program_bench,57.2,https://media.x.ai/v1/website/4p5-5184fdf9.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""official ProgramBench sandbox"",""judge"":""248,000+ fuzz-generated behavioral tests"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""compiled executable and documentation; no internet or decompilation""}",True,model_card,verified,card45 ProgramBench Tests-passed rate (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,single_turn_hallucination_rate,0.98,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""SpaceXAI internal single-turn factuality harness"",""judge"":""separate factual-support grader"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""none""}",True,model_card,verified,card46 Single-turn hallucination Hallucination rate (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,spacexai_mts_accuracy,57.1,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""Grok Build"",""judge"":""task correctness under fixed rollout budget"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""Grok Build coding and research tools""}",True,model_card,verified,card46 SpaceXAI MTS Eval Correct (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,swe_atlas_codebase_qna,84.0,https://media.x.ai/v1/website/4p5-5184fdf9.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""SWE-Atlas-QnA tracked split"",""judge"":""repository-ground-truth rubric"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""codebase reading and exploration""}",True,model_card,verified,card45 SWE-Atlas-QnA Accuracy (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,swe_bench_multilingual,78.0,https://media.x.ai/v1/website/4p5-5184fdf9.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""fixed SWE-bench Multilingual agent scaffold"",""judge"":""language-appropriate isolated tests"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic repository shell/editor""}",True,model_card,verified,card45 SWE-Bench Multilingual Resolve rate (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,swe_marathon_v1_0,29.0,https://media.x.ai/v1/website/4p5-5184fdf9.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""SWE-Marathon v1.0 harness"",""judge"":""multi-layer resolution verification"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic code execution""}",True,model_card,verified,card45 SWE-Marathon v1.0 Resolution rate (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,swe_marathon_v1_1,29.4,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""SWE-Marathon v1.1 harness"",""judge"":""multi-layer resolution verification"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic code execution""}",True,model_card,verified,card46 SWE-Marathon v1.1 Resolution rate (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,tau3_banking,32.6,https://media.x.ai/v1/website/4p5-5184fdf9.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""Artificial Analysis tau3-banking harness"",""judge"":""final-state task accuracy"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""banking APIs and simulated user""}",True,model_card,verified,card45 tau3-banking Accuracy (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,terminal_bench_3_0,15.7,https://x.ai/news/grok-4-6,"{""context"":""unknown"",""effort"":""high"",""harness"":""Terminal-Bench 3.0 / FrontierBench fixed harness"",""judge"":""verified task-success evaluator"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""container terminal""}",False,official_blog,verified,release46 Terminal-Bench v3.0 Task success rate (%); source effort=high; benchmark protocol matches canonical=False. Unknown source fields remain unknown.
grok-4.5,vals_index,65.3,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""Vals Index live harness"",""judge"":""Vals weighted suite score"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""benchmark-specified professional tools""}",True,model_card,verified,card46 Vals Index Suite score (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,vct_accuracy,65.5,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""VCT capability harness"",""judge"":""exact VCT accuracy"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""multimodal input; no safeguards""}",True,model_card,verified,card46 Virology Capabilities Test Accuracy (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,wmdp_bio_accuracy,90.9,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""WMDP official evaluation"",""judge"":""multiple-choice exact accuracy"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""none; no safeguards""}",True,model_card,verified,card46 WMDP WMDP-Bio accuracy (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,wmdp_chem_accuracy,87.3,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""WMDP official evaluation"",""judge"":""multiple-choice exact accuracy"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""none; no safeguards""}",True,model_card,verified,card46 WMDP WMDP-Chem accuracy (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,xai_bio_refusal_accuracy,97.9,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""SpaceXAI Autointent-Bio suite"",""judge"":""dangerous-query refusal accuracy"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""full safeguards""}",True,model_card,verified,card46 CBRN refusals Bio refusal accuracy (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,xai_chem_refusal_accuracy,96.7,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""SpaceXAI Autointent-Chem suite"",""judge"":""dangerous-query refusal accuracy"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""full safeguards""}",True,model_card,verified,card46 CBRN refusals Chem refusal accuracy (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,xai_child_safety_compliance,0.0,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""SpaceXAI child-safety suite"",""judge"":""CSAM and child-safety compliance grader"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""standard safeguards""}",True,model_card,verified,card46 Child safety Compliance (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,xai_epistemic_bias_rate,20.4,https://media.x.ai/v1/website/4p5-5184fdf9.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""SpaceXAI epistemic-bias suite"",""judge"":""opposing-framing bias grader"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""none""}",True,model_card,verified,card45 Epistemic Bias Bias rate (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,xai_general_refusal_compliance,1.1,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""SpaceXAI general-refusal suite"",""judge"":""multilingual should-refuse compliance grader"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""standard safeguards""}",True,model_card,verified,card46 General refusals Compliance (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,xai_inferenceeval_accuracy,41.3,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""SpaceXAI InferenceEval harness"",""judge"":""hidden GPU unit score gated by integration probe"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""Grok Build coding tools and GPU tests""}",True,model_card,verified,card46 InferenceEval Accuracy (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,xai_kernelbenchinternal_accuracy_efficiency,58.4,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""SpaceXAI KernelBenchInternal harness"",""judge"":""correctness and measured speedup composite"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""sandboxed GPU coding tools""}",True,model_card,verified,card46 KernelBenchInternal Accuracy / efficiency (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,xai_mask_rectified_dishonesty,0.67,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""SpaceXAI MASK-Rectified harness"",""judge"":""MASK-Rectified dishonesty grader"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""none""}",True,model_card,verified,card46 MASK-Rectified Dishonesty (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,xai_rn_refusal_accuracy,97.9,https://media.x.ai/v1/website/4p5-5184fdf9.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""FORTRESS-RN evaluation"",""judge"":""dangerous-query refusal accuracy"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""full safeguards""}",True,model_card,verified,card45 CBRN refusals R/N refusal accuracy (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,xai_securecodereview_reward,49.4,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""SpaceXAI SecureCodeReview harness"",""judge"":""security-fix reward with regression checks"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""fixed secure-code-review tools""}",True,model_card,verified,card46 SecureCodeReview Reward (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,xai_self_harm_compliance,0.5,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""SpaceXAI self-harm suite"",""judge"":""self-harm assistance or failed-redirection compliance grader"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""standard safeguards""}",True,model_card,verified,card46 Self-harm refusals Compliance (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,xai_standard_jailbreak_compliance,0.73,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""SpaceXAI jailbreak suite"",""judge"":""should-refuse compliance grader"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""standard safeguards""}",True,model_card,verified,card46 Jailbreaks Standard jailbreak compliance (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.5,xai_sycophancy_rate,0.01,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""SpaceXAI internal sycophancy suite"",""judge"":""accuracy-drop sycophancy grader"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""none""}",True,model_card,verified,card46 Sycophancy Sycophancy (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,aa_briefcase_elo,1577.0,https://x.ai/news/grok-4-6,"{""context"":""unknown"",""effort"":""high"",""harness"":""Artificial Analysis AA-Briefcase harness"",""judge"":""rubric plus analytical and presentation pairwise Elo"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""offline multi-file knowledge-work tools""}",True,official_blog,verified,release46 AA-Briefcase Elo; source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,aa_intelligence_index_v4_1,61.0,https://x.ai/news/grok-4-6,"{""context"":""unknown"",""effort"":""high"",""harness"":""Artificial Analysis live v4.1 harness"",""judge"":""Artificial Analysis weighted composite"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""benchmark-specified""}",True,official_blog,verified,release46 AA Intelligence Index Composite score; source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,apex_agents,57.5,https://x.ai/news/grok-4-6,"{""context"":""unknown"",""effort"":""high"",""harness"":""Mercor APEX-Agents harness"",""judge"":""expert all-criteria-pass rubric"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""multi-application professional tools""}",True,official_blog,verified,release46 APEX-Agents Pass@1 (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,apex_swe,56.4,https://x.ai/news/grok-4-6,"{""context"":""unknown"",""effort"":""high"",""harness"":""Mercor APEX-SWE harness"",""judge"":""Mercor integration and observability task verifier"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic software-engineering environment""}",True,official_blog,verified,release46 APEX-SWE Pass@1 (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,biosecurity_vct_accuracy,47.8,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""SpaceXAI internal Biosecurity VCT harness"",""judge"":""overall should-refuse and should-comply accuracy"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""no safeguards""}",True,model_card,verified,card46 Biosecurity VCT Accuracy (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,biousebench_severity5_refusal_rate,90.7,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""SpaceXAI internal BioUseBench harness"",""judge"":""severity-5 refusal grader"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""standard safeguards""}",True,model_card,verified,card46 BioUseBench Severity-5 refusal rate (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,cadgenbench_generation_cad_score,40.9,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""CadGenBench generation harness"",""judge"":""generation CAD score"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""CAD generation agent tools""}",True,model_card,verified,card46 CadGenBench Generation Reward (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,cve_bench_reward,39.8,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""CVE-Bench sandbox"",""judge"":""CVE-Bench exploit reward"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unrestricted sandboxed web-application tools""}",True,model_card,verified,card46 CVE-Bench Reward (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,cybergym,79.7,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""CyberGym unrestricted capability harness"",""judge"":""PoC reproduced on vulnerable but not fixed version"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""unrestricted agentic code execution""}",True,model_card,verified,card46 CyberGym Mean Reproduced (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,deep_swe_v1_1,65.9,https://x.ai/news/grok-4-6,"{""context"":""unknown"",""effort"":""high"",""harness"":""mini-swe-agent run by Datacurve"",""judge"":""isolated repository verifier"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic repository shell/editor""}",True,official_blog,verified,release46 DeepSWE v1.1 Pass@1 (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,deepsearchqa_acc,81.6,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""SpaceXAI internal DeepSearchQA implementation"",""judge"":""DeepSearchQA answer-accuracy grader"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""web search agent""}",False,model_card,verified,card46 DeepSearchQA Accuracy (%); source effort=high; benchmark protocol matches canonical=False. Unknown source fields remain unknown.
grok-4.6,eebench_v1_core_reward,53.0,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""EEBench v1 core public leaderboard harness"",""judge"":""physical correctness and functionality reward"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""electrical-engineering design tools""}",True,model_card,verified,card46 EEBench v1 core Reward (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,frontiercode_extended_v1_1,61.3,https://x.ai/news/grok-4-6,"{""context"":""unknown"",""effort"":""high"",""harness"":""FrontierCode v1.1 Extended harness"",""judge"":""maintainer-authored rubrics, verifiers, and hard blockers"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""fixed coding-agent scaffold""}",True,official_blog,verified,release46 FrontierCode v1.1 Extended Extended score (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,gdpval_aa_v2_elo,1753.0,https://x.ai/news/grok-4-6,"{""context"":""unknown"",""effort"":""high"",""harness"":""Artificial Analysis GDPval-AA v2 harness"",""judge"":""pairwise Bradley-Terry Elo"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""benchmark-specified knowledge-work tools""}",True,official_blog,verified,release46 GDPVal-AA v2 Elo; source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,hackerbench_v0_2_benign_refusal,0.8,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""HackerBench v0.2 with standard safeguards"",""judge"":""benign refusal grader"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""constrained cyber task tools""}",True,model_card,verified,card46 HackerBench v0.2 Benign Refusal (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,hackerbench_v0_2_harmful_compliance,16.7,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""HackerBench v0.2 with standard safeguards"",""judge"":""harmful or dual-use compliance grader"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""constrained cyber task tools""}",True,model_card,verified,card46 HackerBench v0.2 Harmful/Dual-Use Compliance (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,harvey_lab_vals_final_score,15.8,https://x.ai/news/grok-4-6,"{""context"":""unknown"",""effort"":""high"",""harness"":""Vals implementation of Harvey LAB"",""judge"":""expert all-pass final score"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""Vals legal-agent file tools""}",True,official_blog,verified,release46 Harvey LAB (Vals) Final score (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,lab_bench_protocolqa_accuracy,79.6,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""LAB-Bench ProtocolQA harness"",""judge"":""open-ended single-error accuracy grader"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""none; no safeguards""}",False,model_card,verified,card46 ProtocolQA Open-Ended Accuracy (%); source effort=high; benchmark protocol matches canonical=False. Unknown source fields remain unknown.
grok-4.6,officeqa_pro,63.2,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""OfficeQA Pro official harness"",""judge"":""OfficeQA Pro exact-match accuracy"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""document and rendered-image analysis""}",True,model_card,verified,card46 OfficeQA Pro Accuracy (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,parametric_cad_bench_combined,87.8,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""gNucleus Parametric CAD Bench harness"",""judge"":""combined executability and geometric reward"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""parametric CAD agent tools""}",True,model_card,verified,card46 Parametric CAD Bench Reward (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,single_turn_hallucination_rate,1.7,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""SpaceXAI internal single-turn factuality harness"",""judge"":""separate factual-support grader"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""none""}",True,model_card,verified,card46 Single-turn hallucination Hallucination rate (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,spacexai_mts_accuracy,61.1,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""Grok Build"",""judge"":""task correctness under fixed rollout budget"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""Grok Build coding and research tools""}",True,model_card,verified,card46 SpaceXAI MTS Eval Correct (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,strongreject_compliance,3.9,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""StrongREJECT evaluation"",""judge"":""StrongREJECT compliance grader"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""standard safeguards""}",True,model_card,verified,card46 Jailbreaks StrongREJECT compliance (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,swe_marathon_v1_1,31.9,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""SWE-Marathon v1.1 harness"",""judge"":""multi-layer resolution verification"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic code execution""}",True,model_card,verified,card46 SWE-Marathon v1.1 Resolution rate (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,terminal_bench_3_0,26.0,https://x.ai/news/grok-4-6,"{""context"":""unknown"",""effort"":""high"",""harness"":""Terminal-Bench 3.0 / FrontierBench fixed harness"",""judge"":""verified task-success evaluator"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""container terminal""}",False,official_blog,verified,release46 Terminal-Bench v3.0 Task success rate (%); source effort=high; benchmark protocol matches canonical=False. Unknown source fields remain unknown.
grok-4.6,vals_index,71.1,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""Vals Index live harness"",""judge"":""Vals weighted suite score"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""benchmark-specified professional tools""}",True,model_card,verified,card46 Vals Index Suite score (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,vct_accuracy,67.4,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""VCT capability harness"",""judge"":""exact VCT accuracy"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""multimodal input; no safeguards""}",True,model_card,verified,card46 Virology Capabilities Test Accuracy (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,wmdp_bio_accuracy,90.0,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""WMDP official evaluation"",""judge"":""multiple-choice exact accuracy"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""none; no safeguards""}",True,model_card,verified,card46 WMDP WMDP-Bio accuracy (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,wmdp_chem_accuracy,85.3,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""WMDP official evaluation"",""judge"":""multiple-choice exact accuracy"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""none; no safeguards""}",True,model_card,verified,card46 WMDP WMDP-Chem accuracy (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,wmdp_cyber_accuracy,90.1,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""WMDP official evaluation"",""judge"":""multiple-choice exact accuracy"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""none; no safeguards""}",True,model_card,verified,card46 WMDP WMDP-Cyber accuracy (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,xai_bio_refusal_accuracy,100.0,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""SpaceXAI Autointent-Bio suite"",""judge"":""dangerous-query refusal accuracy"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""full safeguards""}",True,model_card,verified,card46 CBRN refusals Bio refusal accuracy (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,xai_chem_refusal_accuracy,100.0,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""SpaceXAI Autointent-Chem suite"",""judge"":""dangerous-query refusal accuracy"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""full safeguards""}",True,model_card,verified,card46 CBRN refusals Chem refusal accuracy (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,xai_child_safety_compliance,0.0,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""SpaceXAI child-safety suite"",""judge"":""CSAM and child-safety compliance grader"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""standard safeguards""}",True,model_card,verified,card46 Child safety Compliance (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,xai_general_refusal_compliance,0.93,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""SpaceXAI general-refusal suite"",""judge"":""multilingual should-refuse compliance grader"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""standard safeguards""}",True,model_card,verified,card46 General refusals Compliance (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,xai_inferenceeval_accuracy,46.9,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""SpaceXAI InferenceEval harness"",""judge"":""hidden GPU unit score gated by integration probe"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""Grok Build coding tools and GPU tests""}",True,model_card,verified,card46 InferenceEval Accuracy (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,xai_kernelbenchinternal_accuracy_efficiency,70.6,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""SpaceXAI KernelBenchInternal harness"",""judge"":""correctness and measured speedup composite"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""sandboxed GPU coding tools""}",True,model_card,verified,card46 KernelBenchInternal Accuracy / efficiency (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,xai_long_horizon_jailbreak_compliance,1.0,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""SpaceXAI long-horizon jailbreak suite"",""judge"":""long-horizon should-refuse compliance grader"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""standard safeguards""}",True,model_card,verified,card46 Jailbreaks Long-horizon compliance (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,xai_mask_rectified_dishonesty,3.8,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""SpaceXAI MASK-Rectified harness"",""judge"":""MASK-Rectified dishonesty grader"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""none""}",True,model_card,verified,card46 MASK-Rectified Dishonesty (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,xai_securecodereview_reward,58.7,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""SpaceXAI SecureCodeReview harness"",""judge"":""security-fix reward with regression checks"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""fixed secure-code-review tools""}",True,model_card,verified,card46 SecureCodeReview Reward (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,xai_self_harm_compliance,3.7,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""SpaceXAI self-harm suite"",""judge"":""self-harm assistance or failed-redirection compliance grader"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""standard safeguards""}",True,model_card,verified,card46 Self-harm refusals Compliance (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,xai_standard_jailbreak_compliance,0.04,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""SpaceXAI jailbreak suite"",""judge"":""should-refuse compliance grader"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""standard safeguards""}",True,model_card,verified,card46 Jailbreaks Standard jailbreak compliance (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
grok-4.6,xai_sycophancy_rate,0.04,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""high"",""harness"":""SpaceXAI internal sycophancy suite"",""judge"":""accuracy-drop sycophancy grader"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""none""}",True,model_card,verified,card46 Sycophancy Sycophancy (%); source effort=high; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
kimi-k3,apex_swe,48.0,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""Mercor APEX-SWE harness"",""judge"":""Mercor integration and observability task verifier"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic software-engineering environment""}",True,model_card,verified,card46 APEX-SWE Pass@1 (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
kimi-k3,eebench_v1_core_reward,38.3,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""EEBench v1 core public leaderboard harness"",""judge"":""physical correctness and functionality reward"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""electrical-engineering design tools""}",True,model_card,verified,card46 EEBench v1 core Reward (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
kimi-k3,gdpval_aa_v2_elo,1682.0,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""Artificial Analysis GDPval-AA v2 harness"",""judge"":""pairwise Bradley-Terry Elo"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""benchmark-specified knowledge-work tools""}",True,model_card,verified,card46 AA GDPVal v2 AA GDPVal Elo; source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
kimi-k3,swe_marathon_v1_1,48.1,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""SWE-Marathon v1.1 harness"",""judge"":""multi-layer resolution verification"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic code execution""}",True,model_card,verified,card46 SWE-Marathon v1.1 Resolution rate (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
kimi-k3,xai_inferenceeval_accuracy,39.8,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""SpaceXAI InferenceEval harness"",""judge"":""hidden GPU unit score gated by integration probe"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""Grok Build coding tools and GPU tests""}",True,model_card,verified,card46 InferenceEval Accuracy (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
kimi-k3,xai_kernelbenchinternal_accuracy_efficiency,68.5,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""max"",""harness"":""SpaceXAI KernelBenchInternal harness"",""judge"":""correctness and measured speedup composite"",""mode"":""thinking"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""sandboxed GPU coding tools""}",True,model_card,verified,card46 KernelBenchInternal Accuracy / efficiency (%); source effort=max; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
swe-1.7,frontiercode_extended_v1_1,54.3,https://media.x.ai/v1/website/card-4p6-4cd2dc57.pdf,"{""context"":""unknown"",""effort"":""unknown"",""harness"":""FrontierCode v1.1 Extended harness"",""judge"":""maintainer-authored rubrics, verifiers, and hard blockers"",""mode"":""agentic/reasoning"",""prompt_style"":""unknown"",""sampling"":""unknown"",""temperature"":""unknown"",""tools"":""fixed coding-agent scaffold""}",False,model_card,verified,card46 FrontierCode v1.1 Extended Extended score (%); source effort=unknown; benchmark protocol matches canonical=True. Unknown source fields remain unknown.
claude-fable-5,code_arena_frontend_elo,1654.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""high"",""harness"":""Arena Code leaderboard"",""judge"":""human preference"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""107,962-vote Arena snapshot"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Code Arena: Frontend snapshot (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4xsvz.jpg). Historical leaderboard snapshot reproduced by ByteDance. Fable 5 was marked not currently sampled.
claude-opus-4.6,code_arena_frontend_elo,1542.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""Arena Code leaderboard"",""judge"":""human preference"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""107,962-vote Arena snapshot"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Code Arena: Frontend snapshot (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4xsvz.jpg). Historical leaderboard snapshot reproduced by ByteDance.
claude-opus-4.6,officeqa_pro,48.12,https://arxiv.org/abs/2603.08655,"{""effort"":""unknown"",""harness"":""OfficeQA Pro; corpus=full; format=PDF"",""judge"":""0% tolerance exact match"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Claude Code CLI""}",False,leaderboard,verified,Model card Figure 17 (p30). Baseline reproduced from OfficeQA Pro paper section 3.2.
claude-opus-4.7,aethercode,52.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 11 (p53).
claude-opus-4.7,agent_startup_bench,62.3,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""expert review"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""research and professional-deliverable tools""}",False,official_blog,verified,Release blog EN: General-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4u89c.png).
claude-opus-4.7,agents_last_exam,18.4,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""computer-use environment""}",False,official_blog,verified,"Release blog EN: General-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4u89c.png). Turbo is a source dash, not zero."
claude-opus-4.7,agents_last_exam_average_score,40.5,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""computer-use environment""}",False,official_blog,verified,"Release blog EN: General-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4u89c.png). Turbo is a source dash, not zero."
claude-opus-4.7,apex_agents,33.9,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""long-horizon professional-task environment""}",False,official_blog,verified,Release blog EN: General-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4u89c.png).
claude-opus-4.7,babyvision,22.2,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Vision capability chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4yer3.png).
claude-opus-4.7,beyond_aime,79.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 11 (p53).
claude-opus-4.7,blink,70.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
claude-opus-4.7,chartqapro,65.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
claude-opus-4.7,charxiv_descriptive,93.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
claude-opus-4.7,claw_eval_multimodal_pass3,44.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""Pass^3"",""temperature"":""unknown"",""tools"":""high-resolution multimodal agent harness""}",False,official_blog,verified,Release blog EN: Multimodal-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4v1ol.png).
claude-opus-4.7,code_arena_frontend_elo,1563.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""Arena Code leaderboard"",""judge"":""human preference"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""107,962-vote Arena snapshot"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Code Arena: Frontend snapshot (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4xsvz.jpg). Historical leaderboard snapshot reproduced by ByteDance.
claude-opus-4.7,creativework,28.3,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""GUI + MCP in Notion, Canva, and Figma""}",False,official_blog,verified,Release blog EN: Computer-use chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4w8x3.png).
claude-opus-4.7,deep_swe_v1_1,54.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",False,official_blog,verified,Release blog EN: Coding chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4wlbt.png).
claude-opus-4.7,doubao_multi_turn_bench,49.8,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Daily-life/tool-use chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4ujsp.png).
claude-opus-4.7,dude,84.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
claude-opus-4.7,dynamath,63.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
claude-opus-4.7,embspatial_bench,77.2,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Spatial/long-context chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4z2l5.png).
claude-opus-4.7,emma,68.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
claude-opus-4.7,erqa,52.5,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Spatial/long-context chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4z2l5.png).
claude-opus-4.7,frontier_science_research,20.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""research-oriented agent environment""}",False,official_blog,verified,Release blog EN: Frontier-research chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq52cdr.png).
claude-opus-4.7,frontiersci_olympiad,69.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Language/search chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq510uc.png).
claude-opus-4.7,fs_researcher,20.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 11 (p53).
claude-opus-4.7,gameworld,26.5,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""browser-game GUI actions""}",False,official_blog,verified,Release blog EN: Computer-use chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4w8x3.png).
claude-opus-4.7,gdpval_seed_reported_score,82.7,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""source-specific evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""document/spreadsheet deliverable workflow""}",False,official_blog,verified,Release blog EN: General-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4u89c.png).
claude-opus-4.7,hle_tools_text,54.7,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""search""}",False,official_blog,verified,Release blog EN: Language/search chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq510uc.png).
claude-opus-4.7,hle_verified,46.9,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Language/search chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq510uc.png).
claude-opus-4.7,horizon_math_pass12,4.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""research-oriented agent environment""}",False,official_blog,verified,Release blog EN: Frontier-research chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq52cdr.png).
claude-opus-4.7,image2floorplan_avg_score,50.2,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""high-resolution multimodal agent harness""}",False,official_blog,verified,Release blog EN: Multimodal-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4v1ol.png).
claude-opus-4.7,kina,46.7,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Language/search chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq510uc.png).
claude-opus-4.7,live_mathematician_bench_2026_06,36.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 11 (p53).
claude-opus-4.7,mathverse_vision_only,77.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
claude-opus-4.7,mathvision,83.1,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Vision capability chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4yer3.png).
claude-opus-4.7,mathvista,84.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
claude-opus-4.7,measurebench,29.7,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Vision capability chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4yer3.png).
claude-opus-4.7,mmsibench_circular,17.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
claude-opus-4.7,mobileworld,57.1,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""mobile GUI actions""}",False,official_blog,verified,Release blog EN: Computer-use chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4w8x3.png).
claude-opus-4.7,msqa,42.8,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Language/search chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq510uc.png).
claude-opus-4.7,nl2repo_bench,58.2,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",False,official_blog,verified,Release blog EN: Coding chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4wlbt.png).
claude-opus-4.7,ocrbench_v2,56.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
claude-opus-4.7,one_million_bench,73.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic professional-task environment""}",False,official_blog,verified,Release blog EN: General-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4u89c.png).
claude-opus-4.7,presentbench,61.8,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""fine-grained rubric evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""slide-generation environment""}",False,official_blog,verified,Release blog EN: General-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4u89c.png).
claude-opus-4.7,program_bench,52.1,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",False,official_blog,verified,Release blog EN: Coding chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4wlbt.png).
claude-opus-4.7,program_bench_almost_resolved_95,2.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official ProgramBench sandbox"",""judge"":""248,000+ behavioral tests"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""compiled executable and documentation""}",False,tech_report,verified,Model card Table 5 (p19).
claude-opus-4.7,program_bench_fully_resolved,0.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official ProgramBench sandbox"",""judge"":""248,000+ behavioral tests"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""compiled executable and documentation""}",False,tech_report,verified,Model card Table 5 (p19).
claude-opus-4.7,realworldqa,75.6,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Vision capability chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4yer3.png).
claude-opus-4.7,scicode,56.4,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Language/search chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq510uc.png).
claude-opus-4.7,seed21_trae_acceptable_delivery_rate,92.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""multidimensional human rating"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""real-repository Trae coding environment""}",False,tech_report,verified,"Model card textual metrics (pp15-16, p32). Acceptable delivery rate."
claude-opus-4.7,seed21_trae_boundary_adherence,4.38,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""multidimensional human rating"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""real-repository Trae coding environment""}",False,tech_report,verified,"Model card textual metrics (pp15-16, p32). Boundary-adherence rating."
claude-opus-4.7,seed21_trae_delivery_completeness,3.83,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""multidimensional human rating"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""real-repository Trae coding environment""}",False,tech_report,verified,"Model card textual metrics (pp15-16, p32). Delivery-completeness score."
claude-opus-4.7,seed21_trae_fully_correct_rate,20.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""multidimensional human rating"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""real-repository Trae coding environment""}",False,tech_report,verified,"Model card textual metrics (pp15-16, p32). Fully correct, ready-to-use artifact rate."
claude-opus-4.7,seed21_trae_fully_usable_rate,54.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""multidimensional human rating"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""real-repository Trae coding environment""}",False,tech_report,verified,"Model card textual metrics (pp15-16, p32). Top fully usable tier."
claude-opus-4.7,seed21_trae_instruction_following,4.11,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""multidimensional human rating"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""real-repository Trae coding environment""}",False,tech_report,verified,"Model card textual metrics (pp15-16, p32). Instruction-following rating."
claude-opus-4.7,seed21_trae_mean_rating,3.967,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""multidimensional human rating"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""real-repository Trae coding environment""}",False,tech_report,verified,"Model card textual metrics (pp15-16, p32). Mean across six rating dimensions."
claude-opus-4.7,seed21_trae_severely_broken_rate,2.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""multidimensional human rating"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""real-repository Trae coding environment""}",False,tech_report,verified,"Model card textual metrics (pp15-16, p32). Severely broken delivery rate; lower is better."
claude-opus-4.7,seed21_trae_unusable_rate,7.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""multidimensional human rating"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""real-repository Trae coding environment""}",False,tech_report,verified,"Model card textual metrics (pp15-16, p32). Unusable tier; lower is better."
claude-opus-4.7,seedclawbench,64.1,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""OpenClaw-style tools and skills""}",False,official_blog,verified,Release blog EN: Daily-life/tool-use chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4ujsp.png).
claude-opus-4.7,seedkernelbench_avg_speedup,10.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",False,tech_report,verified,Model card Table 5 (p19). Source dashes are not zero.
claude-opus-4.7,simplevqa,56.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
claude-opus-4.7,superchem,55.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 11 (p53).
claude-opus-4.7,supergpqa,68.5,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Language/search chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq510uc.png).
claude-opus-4.7,swe_atlas,38.7,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",False,official_blog,verified,Release blog EN: Coding chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4wlbt.png).
claude-opus-4.7,trae_artifacts,53.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",False,tech_report,verified,Model card Table 5 (p19).
claude-opus-4.7,trae_code_gen_js,66.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",False,tech_report,verified,Model card Table 5 (p19).
claude-opus-4.7,trae_code_gen_python,83.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",False,tech_report,verified,Model card Table 5 (p19).
claude-opus-4.7,trae_error_fix_go,66.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",False,tech_report,verified,Model card Table 5 (p19).
claude-opus-4.7,trae_error_fix_java,73.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",False,tech_report,verified,Model card Table 5 (p19).
claude-opus-4.7,trae_error_fix_js,78.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",False,tech_report,verified,Model card Table 5 (p19).
claude-opus-4.7,trae_error_fix_python,76.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",False,tech_report,verified,Model card Table 5 (p19).
claude-opus-4.7,trae_repo_env,63.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",False,tech_report,verified,Model card Table 5 (p19).
claude-opus-4.7,treebench,36.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
claude-opus-4.7,visfactor,30.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
claude-opus-4.7,visulogic,32.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
claude-opus-4.7,vlms_are_biased,30.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
claude-opus-4.7,webbench,75.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",False,tech_report,verified,Model card Table 5 (p19).
claude-opus-4.7,wildclaw_bench_60_openclaw,67.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""high-resolution multimodal agent harness""}",False,official_blog,verified,Release blog EN: Multimodal-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4v1ol.png).
claude-opus-4.7,workspace_bench_openclaw_100,55.1,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""100-task OpenClaw setting"",""judge"":""rubric-based evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""OpenClaw workspace tools""}",False,official_blog,verified,Release blog EN: General-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4u89c.png).
claude-opus-4.7,workspace_bench_openclaw_100_pass100,18.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""100-task OpenClaw setting"",""judge"":""rubric-based evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""OpenClaw workspace tools""}",False,tech_report,verified,Model card Table 1 (p9).
claude-opus-4.7,workspace_bench_openclaw_100_pass30,76.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""100-task OpenClaw setting"",""judge"":""rubric-based evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""OpenClaw workspace tools""}",False,tech_report,verified,Model card Table 1 (p9).
claude-opus-4.7,workspace_bench_openclaw_100_pass50,56.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""100-task OpenClaw setting"",""judge"":""rubric-based evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""OpenClaw workspace tools""}",False,tech_report,verified,Model card Table 1 (p9).
claude-opus-4.7,workspace_bench_openclaw_100_pass70,40.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""100-task OpenClaw setting"",""judge"":""rubric-based evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""OpenClaw workspace tools""}",False,tech_report,verified,Model card Table 1 (p9).
claude-opus-4.7,workspace_bench_openclaw_100_pass90,20.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""100-task OpenClaw setting"",""judge"":""rubric-based evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""OpenClaw workspace tools""}",False,tech_report,verified,Model card Table 1 (p9).
claude-opus-4.7,worldbench,52.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
claude-opus-4.7,worldvqa,35.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
claude-opus-4.7,xdailybench,69.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Daily-life/tool-use chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4ujsp.png).
claude-opus-4.7,zerobench_main,8.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Vision capability chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4yer3.png).
claude-opus-4.7,zerobench_sub,37.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
claude-opus-4.8,code_arena_frontend_elo,1565.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""Arena Code leaderboard"",""judge"":""human preference"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""107,962-vote Arena snapshot"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Code Arena: Frontend snapshot (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4xsvz.jpg). Historical leaderboard snapshot reproduced by ByteDance.
claude-opus-4.8,seed21_dirtyfilter_precision,94.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""300M-token budget; identical validation set"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""rule-only evaluator/optimizer loop""}",False,tech_report,verified,Model card Figure 19 (p32).
claude-opus-4.8,seed21_dirtyfilter_validation_recall,51.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""300M-token budget; identical validation set"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""rule-only evaluator/optimizer loop""}",False,tech_report,verified,Model card Figure 19 (p32).
claude-sonnet-4.6,code_arena_frontend_elo,1521.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""Arena Code leaderboard"",""judge"":""human preference"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""107,962-vote Arena snapshot"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Code Arena: Frontend snapshot (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4xsvz.jpg). Historical leaderboard snapshot reproduced by ByteDance.
doubao-seed-2.1-deep-think,frontier_science_research,40.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""reason -> verify -> revise -> select loop"",""judge"":""benchmark-specified"",""mode"":""deep-think"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""web search + sandboxed code execution""}",False,tech_report,verified,Model card Table 7 (p24).
doubao-seed-2.1-deep-think,imo_2025,81.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""reason -> verify -> revise -> select loop"",""judge"":""benchmark-specified"",""mode"":""deep-think"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""web search + sandboxed code execution""}",True,tech_report,verified,Model card Table 7 (p24).
doubao-seed-2.1-deep-think,imo_proofbench_advanced,83.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""reason -> verify -> revise -> select loop"",""judge"":""benchmark-specified"",""mode"":""deep-think"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""web search + sandboxed code execution""}",True,tech_report,verified,Model card Table 7 (p24).
doubao-seed-2.1-deep-think,ipho_2025_theory,89.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""reason -> verify -> revise -> select loop"",""judge"":""benchmark-specified"",""mode"":""deep-think"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""web search + sandboxed code execution""}",True,tech_report,verified,Model card Table 7 (p24).
doubao-seed-2.1-pro,aethercode,65.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 11 (p53).
doubao-seed-2.1-pro,agent_startup_bench,68.8,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""expert review"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""research and professional-deliverable tools""}",True,official_blog,verified,Release blog EN: General-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4u89c.png).
doubao-seed-2.1-pro,agents_last_exam,19.5,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""computer-use environment""}",True,official_blog,verified,"Release blog EN: General-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4u89c.png). Turbo is a source dash, not zero."
doubao-seed-2.1-pro,agents_last_exam_average_score,41.4,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""computer-use environment""}",True,official_blog,verified,"Release blog EN: General-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4u89c.png). Turbo is a source dash, not zero."
doubao-seed-2.1-pro,arc_agi_2,62.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 11 (p53).
doubao-seed-2.1-pro,babyvision,73.7,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Vision capability chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4yer3.png).
doubao-seed-2.1-pro,beyond_aime,87.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 11 (p53).
doubao-seed-2.1-pro,blink,81.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 9 (p51).
doubao-seed-2.1-pro,chartqapro,70.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 9 (p51).
doubao-seed-2.1-pro,charxiv_descriptive,95.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 9 (p51).
doubao-seed-2.1-pro,charxiv_reasoning,85.4,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Vision capability chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4yer3.png).
doubao-seed-2.1-pro,claw_eval_multimodal_pass3,51.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""Pass^3"",""temperature"":""unknown"",""tools"":""high-resolution multimodal agent harness""}",True,official_blog,verified,Release blog EN: Multimodal-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4v1ol.png).
doubao-seed-2.1-pro,contphy,63.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 10 (p52).
doubao-seed-2.1-pro,creativework,42.5,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""GUI + MCP in Notion, Canva, and Figma""}",True,official_blog,verified,Release blog EN: Computer-use chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4w8x3.png).
doubao-seed-2.1-pro,crossvid,65.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 10 (p52).
doubao-seed-2.1-pro,cybergym,68.7,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""cybersecurity agent environment""}",True,official_blog,verified,Release blog EN: Coding chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4wlbt.png). Release blog reports Pro=68.7; PDF Table 5 reports Pro=70.2.
doubao-seed-2.1-pro,doubao_multi_turn_bench,52.5,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Daily-life/tool-use chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4ujsp.png).
doubao-seed-2.1-pro,dude,82.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 9 (p51).
doubao-seed-2.1-pro,dynamath,73.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 9 (p51).
doubao-seed-2.1-pro,embspatial_bench,83.4,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Spatial/long-context chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4z2l5.png).
doubao-seed-2.1-pro,emma,79.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 9 (p51).
doubao-seed-2.1-pro,erqa,72.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Spatial/long-context chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4z2l5.png).
doubao-seed-2.1-pro,finance_agent,60.7,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""LLM-as-judge rubric and contradiction grader"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""financial-analysis agent tools""}",True,official_blog,verified,Release blog EN: General-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4u89c.png).
doubao-seed-2.1-pro,frontiercs_algorithmic_v1,28.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""research and code-execution environment""}",True,tech_report,verified,Model card Table 6 (p24). Claude Opus 4.7 is a source dash; official sub-score name unresolved.
doubao-seed-2.1-pro,frontiercs_overall_v1,29.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""research and code-execution environment""}",True,tech_report,verified,Model card Table 6 (p24). Claude Opus 4.7 is a source dash; official sub-score name unresolved.
doubao-seed-2.1-pro,frontiercs_research_v1,46.3,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""research and code-execution environment""}",True,official_blog,verified,Release blog EN: Frontier-research chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq52cdr.png). Claude Opus 4.7 is a source dash; official sub-score name unresolved.
doubao-seed-2.1-pro,fs_researcher,28.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 11 (p53).
doubao-seed-2.1-pro,gameworld,31.2,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""browser-game GUI actions""}",True,official_blog,verified,Release blog EN: Computer-use chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4w8x3.png).
doubao-seed-2.1-pro,gdpval_seed_reported_score,87.9,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""source-specific evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""document/spreadsheet deliverable workflow""}",True,official_blog,verified,Release blog EN: General-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4u89c.png).
doubao-seed-2.1-pro,hle_verified,42.9,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Language/search chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq510uc.png).
doubao-seed-2.1-pro,image2floorplan_avg_score,48.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""high-resolution multimodal agent harness""}",True,official_blog,verified,Release blog EN: Multimodal-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4v1ol.png).
doubao-seed-2.1-pro,imo_2025,65.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""Deep Think comparison baseline"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Deep Think section baseline; exact tools undisclosed""}",True,tech_report,verified,Model card Table 7 (p24).
doubao-seed-2.1-pro,imo_proofbench_advanced,54.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""Deep Think comparison baseline"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Deep Think section baseline; exact tools undisclosed""}",True,tech_report,verified,Model card Table 7 (p24).
doubao-seed-2.1-pro,ipho_2025_theory,79.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""Deep Think comparison baseline"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Deep Think section baseline; exact tools undisclosed""}",True,tech_report,verified,Model card Table 7 (p24).
doubao-seed-2.1-pro,kina,48.3,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Language/search chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq510uc.png).
doubao-seed-2.1-pro,live_mathematician_bench_2026_06,20.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 11 (p53).
doubao-seed-2.1-pro,livesports_3k,76.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 10 (p52).
doubao-seed-2.1-pro,longvideobench,80.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 10 (p52).
doubao-seed-2.1-pro,lvbench,78.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Long/streaming-video chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq50ia3.png).
doubao-seed-2.1-pro,mathverse_vision_only,89.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 9 (p51).
doubao-seed-2.1-pro,mathvision,92.6,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Vision capability chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4yer3.png).
doubao-seed-2.1-pro,mathvista,90.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 9 (p51).
doubao-seed-2.1-pro,measurebench,62.9,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Vision capability chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4yer3.png).
doubao-seed-2.1-pro,minerva,70.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 10 (p52). Source marks this row with an unexplained dagger. Source marks this row with an unexplained dagger.
doubao-seed-2.1-pro,mmlongbench,78.3,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Spatial/long-context chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4z2l5.png). Source dashes are not zero.
doubao-seed-2.1-pro,mmmu_pro,81.6,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Vision capability chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4yer3.png).
doubao-seed-2.1-pro,mmsibench_circular,35.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 9 (p51).
doubao-seed-2.1-pro,mobileworld,73.1,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""mobile GUI actions""}",True,official_blog,verified,Release blog EN: Computer-use chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4w8x3.png).
doubao-seed-2.1-pro,motionbench,74.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 10 (p52).
doubao-seed-2.1-pro,msqa,50.2,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Language/search chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq510uc.png).
doubao-seed-2.1-pro,ocrbench_v2,63.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 9 (p51).
doubao-seed-2.1-pro,officeqa_pro,72.2,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""high-resolution multimodal agent harness""}",True,official_blog,verified,Release blog EN: Multimodal-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4v1ol.png).
doubao-seed-2.1-pro,one_million_bench,68.8,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic professional-task environment""}",True,official_blog,verified,Release blog EN: General-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4u89c.png).
doubao-seed-2.1-pro,osworld,78.8,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""GUI + shell/commands + filesystem tools""}",True,official_blog,verified,Release blog EN: Computer-use chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4w8x3.png).
doubao-seed-2.1-pro,ovbench,70.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Long/streaming-video chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq50ia3.png).
doubao-seed-2.1-pro,ovobench,80.7,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Long/streaming-video chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq50ia3.png).
doubao-seed-2.1-pro,posttrain_bench,16.5,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""research-oriented agent environment""}",True,official_blog,verified,Release blog EN: Frontier-research chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq52cdr.png). Source dashes are not zero.
doubao-seed-2.1-pro,presentbench,54.6,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""fine-grained rubric evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""slide-generation environment""}",True,official_blog,verified,Release blog EN: General-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4u89c.png).
doubao-seed-2.1-pro,program_bench,50.3,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",True,official_blog,verified,Release blog EN: Coding chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4wlbt.png).
doubao-seed-2.1-pro,program_bench_almost_resolved_95,1.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official ProgramBench sandbox"",""judge"":""248,000+ behavioral tests"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""compiled executable and documentation""}",True,tech_report,verified,Model card Table 5 (p19).
doubao-seed-2.1-pro,program_bench_fully_resolved,0.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official ProgramBench sandbox"",""judge"":""248,000+ behavioral tests"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""compiled executable and documentation""}",True,tech_report,verified,Model card Table 5 (p19).
doubao-seed-2.1-pro,realworldqa,86.7,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Vision capability chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4yer3.png).
doubao-seed-2.1-pro,scicode,59.8,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Language/search chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq510uc.png).
doubao-seed-2.1-pro,seed21_claudecode_opus46_loss_rate,29.6,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""anonymous developer preference"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""real-repository coding environment""}",True,official_blog,verified,Release blog EN: Crowdsourced coding chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4wx8r.png).
doubao-seed-2.1-pro,seed21_claudecode_opus46_net_win_rate,29.6,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""anonymous developer preference"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""real-repository coding environment""}",True,official_blog,verified,Release blog EN: Crowdsourced coding chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4wx8r.png). Source prints +29.6 percentage points.
doubao-seed-2.1-pro,seed21_claudecode_opus46_tie_rate,11.3,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""anonymous developer preference"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""real-repository coding environment""}",True,official_blog,verified,Release blog EN: Crowdsourced coding chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4wx8r.png).
doubao-seed-2.1-pro,seed21_claudecode_opus46_win_rate,59.1,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""anonymous developer preference"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""real-repository coding environment""}",True,official_blog,verified,Release blog EN: Crowdsourced coding chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4wx8r.png). 136 wins / 26 ties / 68 losses over 230 comparisons.
doubao-seed-2.1-pro,seed21_dirtyfilter_precision,94.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""300M-token budget; identical validation set"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""rule-only evaluator/optimizer loop""}",True,tech_report,verified,Model card Figure 19 (p32).
doubao-seed-2.1-pro,seed21_dirtyfilter_validation_recall,43.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""300M-token budget; identical validation set"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""rule-only evaluator/optimizer loop""}",True,tech_report,verified,Model card Figure 19 (p32).
doubao-seed-2.1-pro,seed21_trae_acceptable_delivery_rate,94.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""multidimensional human rating"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""real-repository Trae coding environment""}",True,tech_report,verified,"Model card textual metrics (pp15-16, p32). Acceptable delivery rate."
doubao-seed-2.1-pro,seed21_trae_boundary_adherence,4.16,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""multidimensional human rating"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""real-repository Trae coding environment""}",True,tech_report,verified,"Model card textual metrics (pp15-16, p32). Boundary-adherence rating."
doubao-seed-2.1-pro,seed21_trae_delivery_completeness,3.97,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""multidimensional human rating"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""real-repository Trae coding environment""}",True,tech_report,verified,"Model card textual metrics (pp15-16, p32). Delivery-completeness score."
doubao-seed-2.1-pro,seed21_trae_fully_correct_rate,29.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""multidimensional human rating"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""real-repository Trae coding environment""}",True,tech_report,verified,"Model card textual metrics (pp15-16, p32). Fully correct, ready-to-use artifact rate."
doubao-seed-2.1-pro,seed21_trae_fully_usable_rate,58.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""multidimensional human rating"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""real-repository Trae coding environment""}",True,tech_report,verified,"Model card textual metrics (pp15-16, p32). Top fully usable tier."
doubao-seed-2.1-pro,seed21_trae_instruction_following,3.96,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""multidimensional human rating"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""real-repository Trae coding environment""}",True,tech_report,verified,"Model card textual metrics (pp15-16, p32). Instruction-following rating."
doubao-seed-2.1-pro,seed21_trae_mean_rating,3.967,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""multidimensional human rating"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""real-repository Trae coding environment""}",True,tech_report,verified,"Model card textual metrics (pp15-16, p32). Mean across six rating dimensions."
doubao-seed-2.1-pro,seed21_trae_opus47_preference_win_rate,48.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""multidimensional human rating"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""real-repository Trae coding environment""}",True,tech_report,verified,"Model card textual metrics (pp15-16, p32). 81 Seed wins and 86 Claude wins over 167 valid head-to-head tasks."
doubao-seed-2.1-pro,seed21_trae_severely_broken_rate,0.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""multidimensional human rating"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""real-repository Trae coding environment""}",True,tech_report,verified,"Model card textual metrics (pp15-16, p32). Severely broken delivery rate; lower is better."
doubao-seed-2.1-pro,seed21_trae_unusable_rate,11.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""multidimensional human rating"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""real-repository Trae coding environment""}",True,tech_report,verified,"Model card textual metrics (pp15-16, p32). Unusable tier; lower is better."
doubao-seed-2.1-pro,seedclawbench,66.6,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""OpenClaw-style tools and skills""}",True,official_blog,verified,Release blog EN: Daily-life/tool-use chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4ujsp.png).
doubao-seed-2.1-pro,seedkernelbench_avg_speedup,9.21,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",True,tech_report,verified,Model card Table 5 (p19). Source dashes are not zero.
doubao-seed-2.1-pro,simplevqa,74.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 9 (p51).
doubao-seed-2.1-pro,supergpqa,70.8,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Language/search chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq510uc.png).
doubao-seed-2.1-pro,swe_atlas,35.2,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",True,official_blog,verified,Release blog EN: Coding chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4wlbt.png).
doubao-seed-2.1-pro,tomato,79.5,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Motion/perception chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4zh7f.png).
doubao-seed-2.1-pro,trae_artifacts,51.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",True,tech_report,verified,Model card Table 5 (p19).
doubao-seed-2.1-pro,trae_code_gen_js,62.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",True,tech_report,verified,Model card Table 5 (p19).
doubao-seed-2.1-pro,trae_code_gen_python,75.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",True,tech_report,verified,Model card Table 5 (p19).
doubao-seed-2.1-pro,trae_error_fix_go,63.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",True,tech_report,verified,Model card Table 5 (p19).
doubao-seed-2.1-pro,trae_error_fix_java,66.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",True,tech_report,verified,Model card Table 5 (p19).
doubao-seed-2.1-pro,trae_error_fix_js,74.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",True,tech_report,verified,Model card Table 5 (p19).
doubao-seed-2.1-pro,trae_error_fix_python,70.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",True,tech_report,verified,Model card Table 5 (p19).
doubao-seed-2.1-pro,trae_repo_env,55.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",True,tech_report,verified,Model card Table 5 (p19).
doubao-seed-2.1-pro,treebench,71.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 9 (p51).
doubao-seed-2.1-pro,tvbench,80.5,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Motion/perception chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4zh7f.png).
doubao-seed-2.1-pro,video_mme,89.2,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Long/streaming-video chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq50ia3.png). Source marks this row with an unexplained dagger. Source marks this row with an unexplained dagger.
doubao-seed-2.1-pro,videoholmes,68.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 10 (p52). Source marks this row with an unexplained dagger. Source marks this row with an unexplained dagger.
doubao-seed-2.1-pro,videosimpleqa,76.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 10 (p52).
doubao-seed-2.1-pro,visfactor,51.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 9 (p51).
doubao-seed-2.1-pro,visulogic,54.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 9 (p51).
doubao-seed-2.1-pro,vlms_are_biased,83.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 9 (p51).
doubao-seed-2.1-pro,webbench,78.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",True,tech_report,verified,Model card Table 5 (p19).
doubao-seed-2.1-pro,wildclaw_bench_60_openclaw,61.7,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""high-resolution multimodal agent harness""}",True,official_blog,verified,Release blog EN: Multimodal-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4v1ol.png).
doubao-seed-2.1-pro,workspace_bench_openclaw_100,53.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""100-task OpenClaw setting"",""judge"":""rubric-based evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""OpenClaw workspace tools""}",True,official_blog,verified,Release blog EN: General-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4u89c.png).
doubao-seed-2.1-pro,workspace_bench_openclaw_100_pass100,20.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""100-task OpenClaw setting"",""judge"":""rubric-based evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""OpenClaw workspace tools""}",True,tech_report,verified,Model card Table 1 (p9).
doubao-seed-2.1-pro,workspace_bench_openclaw_100_pass30,78.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""100-task OpenClaw setting"",""judge"":""rubric-based evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""OpenClaw workspace tools""}",True,tech_report,verified,Model card Table 1 (p9).
doubao-seed-2.1-pro,workspace_bench_openclaw_100_pass50,60.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""100-task OpenClaw setting"",""judge"":""rubric-based evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""OpenClaw workspace tools""}",True,tech_report,verified,Model card Table 1 (p9).
doubao-seed-2.1-pro,workspace_bench_openclaw_100_pass70,38.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""100-task OpenClaw setting"",""judge"":""rubric-based evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""OpenClaw workspace tools""}",True,tech_report,verified,Model card Table 1 (p9).
doubao-seed-2.1-pro,workspace_bench_openclaw_100_pass90,25.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""100-task OpenClaw setting"",""judge"":""rubric-based evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""OpenClaw workspace tools""}",True,tech_report,verified,Model card Table 1 (p9).
doubao-seed-2.1-pro,worldbench,67.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 9 (p51).
doubao-seed-2.1-pro,worldvqa,53.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 9 (p51).
doubao-seed-2.1-pro,xdailybench,61.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Daily-life/tool-use chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4ujsp.png).
doubao-seed-2.1-pro,zerobench_main,18.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Vision capability chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4yer3.png).
doubao-seed-2.1-pro,zerobench_sub,49.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 9 (p51).
doubao-seed-2.1-pro,zerobench_sub_tools,56.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""source tool-augmented setting; exact tools undisclosed""}",True,tech_report,verified,Model card Table 9 (p51). Parenthesized source value.
doubao-seed-2.1-pro,zerobench_tools,22.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""source tool-augmented setting; exact tools undisclosed""}",True,official_blog,verified,Release blog EN: Vision capability chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4yer3.png). Parenthesized source value.
doubao-seed-2.1-pro-preview,code_arena_frontend_elo,1539.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""Arena Code leaderboard"",""judge"":""human preference"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""107,962-vote Arena snapshot"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Code Arena: Frontend snapshot (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4xsvz.jpg). Historical leaderboard snapshot reproduced by ByteDance.
doubao-seed-2.1-turbo,aethercode,67.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 11 (p53).
doubao-seed-2.1-turbo,agent_startup_bench,54.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""expert review"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""research and professional-deliverable tools""}",True,official_blog,verified,Release blog EN: General-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4u89c.png).
doubao-seed-2.1-turbo,apex_agents,29.2,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""long-horizon professional-task environment""}",True,official_blog,verified,Release blog EN: General-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4u89c.png).
doubao-seed-2.1-turbo,arc_agi_2,61.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 11 (p53).
doubao-seed-2.1-turbo,babyvision,62.9,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Vision capability chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4yer3.png).
doubao-seed-2.1-turbo,beyond_aime,88.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 11 (p53).
doubao-seed-2.1-turbo,blink,79.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 9 (p51).
doubao-seed-2.1-turbo,browsecomp,84.9,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""search""}",True,official_blog,verified,Release blog EN: Language/search chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq510uc.png).
doubao-seed-2.1-turbo,chartqapro,70.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 9 (p51).
doubao-seed-2.1-turbo,charxiv_descriptive,94.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 9 (p51).
doubao-seed-2.1-turbo,charxiv_reasoning,82.5,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Vision capability chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4yer3.png).
doubao-seed-2.1-turbo,claw_eval_multimodal_pass3,46.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""Pass^3"",""temperature"":""unknown"",""tools"":""high-resolution multimodal agent harness""}",True,official_blog,verified,Release blog EN: Multimodal-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4v1ol.png).
doubao-seed-2.1-turbo,contphy,61.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 10 (p52).
doubao-seed-2.1-turbo,creativework,34.5,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""GUI + MCP in Notion, Canva, and Figma""}",True,official_blog,verified,Release blog EN: Computer-use chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4w8x3.png).
doubao-seed-2.1-turbo,crossvid,63.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 10 (p52).
doubao-seed-2.1-turbo,cybergym,67.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""cybersecurity agent environment""}",True,official_blog,verified,Release blog EN: Coding chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4wlbt.png). Release blog reports Pro=68.7; PDF Table 5 reports Pro=70.2.
doubao-seed-2.1-turbo,deep_swe_v1_1,23.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",True,official_blog,verified,Release blog EN: Coding chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4wlbt.png).
doubao-seed-2.1-turbo,doubao_multi_turn_bench,49.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Daily-life/tool-use chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4ujsp.png).
doubao-seed-2.1-turbo,dude,83.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 9 (p51).
doubao-seed-2.1-turbo,dynamath,68.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 9 (p51).
doubao-seed-2.1-turbo,embspatial_bench,82.5,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Spatial/long-context chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4z2l5.png).
doubao-seed-2.1-turbo,emma,78.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 9 (p51).
doubao-seed-2.1-turbo,erqa,71.3,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Spatial/long-context chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4z2l5.png).
doubao-seed-2.1-turbo,finance_agent,56.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""LLM-as-judge rubric and contradiction grader"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""financial-analysis agent tools""}",True,official_blog,verified,Release blog EN: General-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4u89c.png).
doubao-seed-2.1-turbo,frontier_science_research,33.3,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""research-oriented agent environment""}",True,official_blog,verified,Release blog EN: Frontier-research chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq52cdr.png).
doubao-seed-2.1-turbo,frontiercs_algorithmic_v1,33.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""research and code-execution environment""}",True,tech_report,verified,Model card Table 6 (p24). Claude Opus 4.7 is a source dash; official sub-score name unresolved.
doubao-seed-2.1-turbo,frontiercs_overall_v1,33.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""research and code-execution environment""}",True,tech_report,verified,Model card Table 6 (p24). Claude Opus 4.7 is a source dash; official sub-score name unresolved.
doubao-seed-2.1-turbo,frontiercs_research_v1,50.8,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""research and code-execution environment""}",True,official_blog,verified,Release blog EN: Frontier-research chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq52cdr.png). Claude Opus 4.7 is a source dash; official sub-score name unresolved.
doubao-seed-2.1-turbo,frontiersci_olympiad,76.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Language/search chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq510uc.png).
doubao-seed-2.1-turbo,fs_researcher,23.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 11 (p53).
doubao-seed-2.1-turbo,gameworld,25.9,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""browser-game GUI actions""}",True,official_blog,verified,Release blog EN: Computer-use chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4w8x3.png).
doubao-seed-2.1-turbo,gdpval_seed_reported_score,82.7,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""source-specific evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""document/spreadsheet deliverable workflow""}",True,official_blog,verified,Release blog EN: General-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4u89c.png).
doubao-seed-2.1-turbo,hle_tools_text,54.6,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""search""}",True,official_blog,verified,Release blog EN: Language/search chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq510uc.png).
doubao-seed-2.1-turbo,hle_verified,42.4,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Language/search chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq510uc.png).
doubao-seed-2.1-turbo,horizon_math_pass12,2.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""research-oriented agent environment""}",True,official_blog,verified,Release blog EN: Frontier-research chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq52cdr.png).
doubao-seed-2.1-turbo,image2floorplan_avg_score,35.9,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""high-resolution multimodal agent harness""}",True,official_blog,verified,Release blog EN: Multimodal-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4v1ol.png).
doubao-seed-2.1-turbo,kina,46.6,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Language/search chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq510uc.png).
doubao-seed-2.1-turbo,live_mathematician_bench_2026_06,27.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 11 (p53).
doubao-seed-2.1-turbo,livesports_3k,77.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 10 (p52).
doubao-seed-2.1-turbo,longvideobench,80.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 10 (p52).
doubao-seed-2.1-turbo,lvbench,76.8,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Long/streaming-video chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq50ia3.png).
doubao-seed-2.1-turbo,matharena_apex_2025,35.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 11 (p53). Source dash is not zero.
doubao-seed-2.1-turbo,mathverse_vision_only,89.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 9 (p51).
doubao-seed-2.1-turbo,mathvision,90.1,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Vision capability chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4yer3.png).
doubao-seed-2.1-turbo,mathvista,90.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 9 (p51).
doubao-seed-2.1-turbo,mcpatlas,80.3,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""MCP servers in an agent harness""}",True,official_blog,verified,Release blog EN: Daily-life/tool-use chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4ujsp.png).
doubao-seed-2.1-turbo,measurebench,58.9,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Vision capability chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4yer3.png).
doubao-seed-2.1-turbo,minerva,65.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 10 (p52). Source marks this row with an unexplained dagger. Source marks this row with an unexplained dagger.
doubao-seed-2.1-turbo,mmlongbench,76.9,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Spatial/long-context chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4z2l5.png). Source dashes are not zero.
doubao-seed-2.1-turbo,mmmu_pro,80.1,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Vision capability chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4yer3.png).
doubao-seed-2.1-turbo,mmsibench_circular,31.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 9 (p51).
doubao-seed-2.1-turbo,mobileworld,70.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""mobile GUI actions""}",True,official_blog,verified,Release blog EN: Computer-use chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4w8x3.png).
doubao-seed-2.1-turbo,motionbench,74.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 10 (p52).
doubao-seed-2.1-turbo,msqa,42.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Language/search chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq510uc.png).
doubao-seed-2.1-turbo,nl2repo_bench,43.7,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",True,official_blog,verified,Release blog EN: Coding chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4wlbt.png).
doubao-seed-2.1-turbo,ocrbench_v2,62.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 9 (p51).
doubao-seed-2.1-turbo,officeqa_pro,71.1,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""high-resolution multimodal agent harness""}",True,official_blog,verified,Release blog EN: Multimodal-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4v1ol.png).
doubao-seed-2.1-turbo,one_million_bench,66.6,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic professional-task environment""}",True,official_blog,verified,Release blog EN: General-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4u89c.png).
doubao-seed-2.1-turbo,osworld,76.4,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""GUI + shell/commands + filesystem tools""}",True,official_blog,verified,Release blog EN: Computer-use chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4w8x3.png).
doubao-seed-2.1-turbo,ovbench,69.7,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Long/streaming-video chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq50ia3.png).
doubao-seed-2.1-turbo,ovobench,79.2,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Long/streaming-video chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq50ia3.png).
doubao-seed-2.1-turbo,posttrain_bench,18.3,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""research-oriented agent environment""}",True,official_blog,verified,Release blog EN: Frontier-research chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq52cdr.png). Source dashes are not zero.
doubao-seed-2.1-turbo,presentbench,48.3,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""fine-grained rubric evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""slide-generation environment""}",True,official_blog,verified,Release blog EN: General-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4u89c.png).
doubao-seed-2.1-turbo,program_bench,49.4,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",True,official_blog,verified,Release blog EN: Coding chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4wlbt.png).
doubao-seed-2.1-turbo,program_bench_almost_resolved_95,0.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official ProgramBench sandbox"",""judge"":""248,000+ behavioral tests"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""compiled executable and documentation""}",True,tech_report,verified,Model card Table 5 (p19).
doubao-seed-2.1-turbo,program_bench_fully_resolved,0.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official ProgramBench sandbox"",""judge"":""248,000+ behavioral tests"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""compiled executable and documentation""}",True,tech_report,verified,Model card Table 5 (p19).
doubao-seed-2.1-turbo,realworldqa,86.3,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Vision capability chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4yer3.png).
doubao-seed-2.1-turbo,scicode,57.8,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Language/search chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq510uc.png).
doubao-seed-2.1-turbo,seed21_claudecode_glm51_net_win_rate,23.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""anonymous developer preference"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""real-repository coding environment""}",True,tech_report,verified,"Model card textual metrics (pp15-16, p32). 100 wins / 19 ties / 59 losses over 178 comparisons."
doubao-seed-2.1-turbo,seed21_claudecode_glm51_win_rate,56.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""anonymous developer preference"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""real-repository coding environment""}",True,tech_report,verified,"Model card textual metrics (pp15-16, p32). 100 wins / 19 ties / 59 losses over 178 comparisons."
doubao-seed-2.1-turbo,seedclawbench,63.8,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""OpenClaw-style tools and skills""}",True,official_blog,verified,Release blog EN: Daily-life/tool-use chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4ujsp.png).
doubao-seed-2.1-turbo,seedkernelbench_avg_speedup,8.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",True,tech_report,verified,Model card Table 5 (p19). Source dashes are not zero.
doubao-seed-2.1-turbo,simplevqa,71.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 9 (p51).
doubao-seed-2.1-turbo,superchem,56.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 11 (p53).
doubao-seed-2.1-turbo,supergpqa,67.4,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Language/search chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq510uc.png).
doubao-seed-2.1-turbo,swe_atlas,30.6,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",True,official_blog,verified,Release blog EN: Coding chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4wlbt.png).
doubao-seed-2.1-turbo,swe_bench_pro,57.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",True,official_blog,verified,Release blog EN: Coding chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4wlbt.png).
doubao-seed-2.1-turbo,terminal_bench_2_1,67.6,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",True,official_blog,verified,Release blog EN: Coding chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4wlbt.png).
doubao-seed-2.1-turbo,tomato,56.8,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Motion/perception chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4zh7f.png).
doubao-seed-2.1-turbo,toolathlon,49.1,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic tool use""}",True,official_blog,verified,Release blog EN: Daily-life/tool-use chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4ujsp.png).
doubao-seed-2.1-turbo,trae_artifacts,47.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",True,tech_report,verified,Model card Table 5 (p19).
doubao-seed-2.1-turbo,trae_code_gen_js,59.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",True,tech_report,verified,Model card Table 5 (p19).
doubao-seed-2.1-turbo,trae_code_gen_python,73.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",True,tech_report,verified,Model card Table 5 (p19).
doubao-seed-2.1-turbo,trae_error_fix_go,56.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",True,tech_report,verified,Model card Table 5 (p19).
doubao-seed-2.1-turbo,trae_error_fix_java,66.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",True,tech_report,verified,Model card Table 5 (p19).
doubao-seed-2.1-turbo,trae_error_fix_js,69.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",True,tech_report,verified,Model card Table 5 (p19).
doubao-seed-2.1-turbo,trae_error_fix_python,74.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",True,tech_report,verified,Model card Table 5 (p19).
doubao-seed-2.1-turbo,trae_repo_env,46.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",True,tech_report,verified,Model card Table 5 (p19).
doubao-seed-2.1-turbo,treebench,71.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 9 (p51).
doubao-seed-2.1-turbo,tvbench,77.2,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Motion/perception chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4zh7f.png).
doubao-seed-2.1-turbo,video_mme,89.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Long/streaming-video chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq50ia3.png). Source marks this row with an unexplained dagger. Source marks this row with an unexplained dagger.
doubao-seed-2.1-turbo,videoholmes,67.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 10 (p52). Source marks this row with an unexplained dagger. Source marks this row with an unexplained dagger.
doubao-seed-2.1-turbo,videosimpleqa,71.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 10 (p52).
doubao-seed-2.1-turbo,visfactor,43.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 9 (p51).
doubao-seed-2.1-turbo,visulogic,52.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 9 (p51).
doubao-seed-2.1-turbo,vlms_are_biased,68.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 9 (p51).
doubao-seed-2.1-turbo,webbench,73.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",True,tech_report,verified,Model card Table 5 (p19).
doubao-seed-2.1-turbo,wildclaw_bench_60_openclaw,62.8,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""high-resolution multimodal agent harness""}",True,official_blog,verified,Release blog EN: Multimodal-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4v1ol.png).
doubao-seed-2.1-turbo,workspace_bench_openclaw_100,54.7,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""100-task OpenClaw setting"",""judge"":""rubric-based evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""OpenClaw workspace tools""}",True,official_blog,verified,Release blog EN: General-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4u89c.png).
doubao-seed-2.1-turbo,workspace_bench_openclaw_100_pass100,16.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""100-task OpenClaw setting"",""judge"":""rubric-based evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""OpenClaw workspace tools""}",True,tech_report,verified,Model card Table 1 (p9).
doubao-seed-2.1-turbo,workspace_bench_openclaw_100_pass30,74.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""100-task OpenClaw setting"",""judge"":""rubric-based evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""OpenClaw workspace tools""}",True,tech_report,verified,Model card Table 1 (p9).
doubao-seed-2.1-turbo,workspace_bench_openclaw_100_pass50,59.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""100-task OpenClaw setting"",""judge"":""rubric-based evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""OpenClaw workspace tools""}",True,tech_report,verified,Model card Table 1 (p9).
doubao-seed-2.1-turbo,workspace_bench_openclaw_100_pass70,39.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""100-task OpenClaw setting"",""judge"":""rubric-based evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""OpenClaw workspace tools""}",True,tech_report,verified,Model card Table 1 (p9).
doubao-seed-2.1-turbo,workspace_bench_openclaw_100_pass90,20.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""100-task OpenClaw setting"",""judge"":""rubric-based evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""OpenClaw workspace tools""}",True,tech_report,verified,Model card Table 1 (p9).
doubao-seed-2.1-turbo,worldbench,63.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 9 (p51).
doubao-seed-2.1-turbo,worldvqa,48.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 9 (p51).
doubao-seed-2.1-turbo,xdailybench,56.4,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Daily-life/tool-use chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4ujsp.png).
doubao-seed-2.1-turbo,zerobench_main,11.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,official_blog,verified,Release blog EN: Vision capability chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4yer3.png).
doubao-seed-2.1-turbo,zerobench_sub,49.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",True,tech_report,verified,Model card Table 9 (p51).
doubao-seed-2.1-turbo,zerobench_sub_tools,57.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""source tool-augmented setting; exact tools undisclosed""}",True,tech_report,verified,Model card Table 9 (p51). Parenthesized source value.
doubao-seed-2.1-turbo,zerobench_tools,20.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""source tool-augmented setting; exact tools undisclosed""}",True,official_blog,verified,Release blog EN: Vision capability chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4yer3.png). Parenthesized source value.
gemini-3.1-pro,aethercode,74.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 11 (p53).
gemini-3.1-pro,agent_startup_bench,45.7,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""expert review"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""research and professional-deliverable tools""}",False,official_blog,verified,Release blog EN: General-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4u89c.png).
gemini-3.1-pro,agents_last_exam_average_score,32.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""computer-use environment""}",False,official_blog,verified,"Release blog EN: General-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4u89c.png). Turbo is a source dash, not zero."
gemini-3.1-pro,beyond_aime,90.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 11 (p53).
gemini-3.1-pro,blink,79.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
gemini-3.1-pro,charxiv_descriptive,94.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
gemini-3.1-pro,claw_eval_multimodal_pass3,27.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""Pass^3"",""temperature"":""unknown"",""tools"":""high-resolution multimodal agent harness""}",False,official_blog,verified,Release blog EN: Multimodal-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4v1ol.png).
gemini-3.1-pro,contphy,64.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 10 (p52).
gemini-3.1-pro,creativework,27.4,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""GUI + MCP in Notion, Canva, and Figma""}",False,official_blog,verified,Release blog EN: Computer-use chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4w8x3.png).
gemini-3.1-pro,crossvid,48.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 10 (p52).
gemini-3.1-pro,doubao_multi_turn_bench,52.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Daily-life/tool-use chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4ujsp.png).
gemini-3.1-pro,dude,82.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
gemini-3.1-pro,dynamath,72.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
gemini-3.1-pro,embspatial_bench,84.2,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Spatial/long-context chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4z2l5.png).
gemini-3.1-pro,emma,72.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
gemini-3.1-pro,erqa,70.8,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Spatial/long-context chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4z2l5.png).
gemini-3.1-pro,frontiercs_algorithmic_v1,44.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""research and code-execution environment""}",False,tech_report,verified,Model card Table 6 (p24). Claude Opus 4.7 is a source dash; official sub-score name unresolved.
gemini-3.1-pro,frontiercs_overall_v1,43.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""research and code-execution environment""}",False,tech_report,verified,Model card Table 6 (p24). Claude Opus 4.7 is a source dash; official sub-score name unresolved.
gemini-3.1-pro,frontiercs_research_v1,64.4,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""research and code-execution environment""}",False,official_blog,verified,Release blog EN: Frontier-research chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq52cdr.png). Claude Opus 4.7 is a source dash; official sub-score name unresolved.
gemini-3.1-pro,fs_researcher,16.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 11 (p53).
gemini-3.1-pro,gameworld,21.2,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""browser-game GUI actions""}",False,official_blog,verified,Release blog EN: Computer-use chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4w8x3.png).
gemini-3.1-pro,gdpval_seed_reported_score,67.3,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""source-specific evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""document/spreadsheet deliverable workflow""}",False,official_blog,verified,Release blog EN: General-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4u89c.png).
gemini-3.1-pro,hle_verified,48.2,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Language/search chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq510uc.png).
gemini-3.1-pro,image2floorplan_avg_score,55.1,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""high-resolution multimodal agent harness""}",False,official_blog,verified,Release blog EN: Multimodal-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4v1ol.png).
gemini-3.1-pro,imo_proofbench_advanced,49.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""benchmark-specified""}",False,tech_report,verified,Model card Table 7 (p24).
gemini-3.1-pro,ipho_2025_theory,76.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""benchmark-specified""}",False,tech_report,verified,Model card Table 7 (p24).
gemini-3.1-pro,kina,53.2,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Language/search chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq510uc.png).
gemini-3.1-pro,live_mathematician_bench_2026_06,43.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 11 (p53).
gemini-3.1-pro,livesports_3k,89.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 10 (p52).
gemini-3.1-pro,longvideobench,76.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 10 (p52).
gemini-3.1-pro,lvbench,66.2,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Long/streaming-video chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq50ia3.png).
gemini-3.1-pro,mathverse_vision_only,87.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
gemini-3.1-pro,mathvista,90.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
gemini-3.1-pro,measurebench,44.4,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Vision capability chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4yer3.png).
gemini-3.1-pro,minerva,63.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 10 (p52). Source marks this row with an unexplained dagger. Source marks this row with an unexplained dagger.
gemini-3.1-pro,mmlongbench,70.7,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Spatial/long-context chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4z2l5.png). Source dashes are not zero.
gemini-3.1-pro,mmsibench_circular,27.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
gemini-3.1-pro,mobileworld,48.4,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""mobile GUI actions""}",False,official_blog,verified,Release blog EN: Computer-use chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4w8x3.png).
gemini-3.1-pro,motionbench,69.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 10 (p52).
gemini-3.1-pro,msqa,69.7,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Language/search chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq510uc.png).
gemini-3.1-pro,ocrbench_v2,62.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
gemini-3.1-pro,one_million_bench,60.2,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic professional-task environment""}",False,official_blog,verified,Release blog EN: General-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4u89c.png).
gemini-3.1-pro,ovbench,58.8,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Long/streaming-video chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq50ia3.png).
gemini-3.1-pro,ovobench,64.1,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Long/streaming-video chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq50ia3.png).
gemini-3.1-pro,presentbench,52.1,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""fine-grained rubric evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""slide-generation environment""}",False,official_blog,verified,Release blog EN: General-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4u89c.png).
gemini-3.1-pro,program_bench_almost_resolved_95,1.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official ProgramBench sandbox"",""judge"":""248,000+ behavioral tests"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""compiled executable and documentation""}",False,tech_report,verified,Model card Table 5 (p19).
gemini-3.1-pro,program_bench_fully_resolved,0.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official ProgramBench sandbox"",""judge"":""248,000+ behavioral tests"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""compiled executable and documentation""}",False,tech_report,verified,Model card Table 5 (p19).
gemini-3.1-pro,realworldqa,85.4,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Vision capability chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4yer3.png).
gemini-3.1-pro,seedclawbench,57.1,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""OpenClaw-style tools and skills""}",False,official_blog,verified,Release blog EN: Daily-life/tool-use chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4ujsp.png).
gemini-3.1-pro,simplevqa,69.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
gemini-3.1-pro,supergpqa,76.6,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Language/search chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq510uc.png).
gemini-3.1-pro,swe_atlas,23.6,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",False,official_blog,verified,Release blog EN: Coding chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4wlbt.png).
gemini-3.1-pro,tomato,60.4,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Motion/perception chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4zh7f.png).
gemini-3.1-pro,trae_artifacts,48.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",False,tech_report,verified,Model card Table 5 (p19).
gemini-3.1-pro,trae_code_gen_js,65.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",False,tech_report,verified,Model card Table 5 (p19).
gemini-3.1-pro,trae_code_gen_python,71.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",False,tech_report,verified,Model card Table 5 (p19).
gemini-3.1-pro,trae_error_fix_go,60.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",False,tech_report,verified,Model card Table 5 (p19).
gemini-3.1-pro,trae_error_fix_java,62.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",False,tech_report,verified,Model card Table 5 (p19).
gemini-3.1-pro,trae_error_fix_js,77.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",False,tech_report,verified,Model card Table 5 (p19).
gemini-3.1-pro,trae_error_fix_python,75.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",False,tech_report,verified,Model card Table 5 (p19).
gemini-3.1-pro,trae_repo_env,57.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",False,tech_report,verified,Model card Table 5 (p19).
gemini-3.1-pro,treebench,60.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
gemini-3.1-pro,tvbench,71.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Motion/perception chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4zh7f.png).
gemini-3.1-pro,video_mme,86.7,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Long/streaming-video chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq50ia3.png). Source marks this row with an unexplained dagger. Source marks this row with an unexplained dagger.
gemini-3.1-pro,videoholmes,65.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 10 (p52). Source marks this row with an unexplained dagger. Source marks this row with an unexplained dagger.
gemini-3.1-pro,videosimpleqa,70.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 10 (p52).
gemini-3.1-pro,visfactor,54.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
gemini-3.1-pro,visulogic,44.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
gemini-3.1-pro,vlms_are_biased,73.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
gemini-3.1-pro,webbench,71.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",False,tech_report,verified,Model card Table 5 (p19).
gemini-3.1-pro,wildclaw_bench_60_openclaw,61.1,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""high-resolution multimodal agent harness""}",False,official_blog,verified,Release blog EN: Multimodal-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4v1ol.png).
gemini-3.1-pro,workspace_bench_openclaw_100,32.8,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""100-task OpenClaw setting"",""judge"":""rubric-based evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""OpenClaw workspace tools""}",False,official_blog,verified,Release blog EN: General-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4u89c.png).
gemini-3.1-pro,workspace_bench_openclaw_100_pass100,4.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""100-task OpenClaw setting"",""judge"":""rubric-based evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""OpenClaw workspace tools""}",False,tech_report,verified,Model card Table 1 (p9).
gemini-3.1-pro,workspace_bench_openclaw_100_pass30,49.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""100-task OpenClaw setting"",""judge"":""rubric-based evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""OpenClaw workspace tools""}",False,tech_report,verified,Model card Table 1 (p9).
gemini-3.1-pro,workspace_bench_openclaw_100_pass50,29.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""100-task OpenClaw setting"",""judge"":""rubric-based evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""OpenClaw workspace tools""}",False,tech_report,verified,Model card Table 1 (p9).
gemini-3.1-pro,workspace_bench_openclaw_100_pass70,17.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""100-task OpenClaw setting"",""judge"":""rubric-based evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""OpenClaw workspace tools""}",False,tech_report,verified,Model card Table 1 (p9).
gemini-3.1-pro,workspace_bench_openclaw_100_pass90,9.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""100-task OpenClaw setting"",""judge"":""rubric-based evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""OpenClaw workspace tools""}",False,tech_report,verified,Model card Table 1 (p9).
gemini-3.1-pro,worldbench,62.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
gemini-3.1-pro,worldvqa,44.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
gemini-3.1-pro,xdailybench,35.2,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Daily-life/tool-use chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4ujsp.png).
gemini-3.1-pro,zerobench_main,12.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Vision capability chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4yer3.png).
gemini-3.1-pro,zerobench_sub,41.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
gemini-3.1-pro-deep-think-2026-02,frontier_science_research,23.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""deep-think"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""benchmark-specified""}",False,tech_report,verified,Model card Table 7 (p24).
gemini-3.1-pro-deep-think-2026-02,imo_proofbench_advanced,76.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""deep-think"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""benchmark-specified""}",False,tech_report,verified,Model card Table 7 (p24).
gemini-3.5-flash,code_arena_frontend_elo,1506.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""Arena Code leaderboard"",""judge"":""human preference"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""107,962-vote Arena snapshot"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Code Arena: Frontend snapshot (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4xsvz.jpg). Historical leaderboard snapshot reproduced by ByteDance.
gemini-3.5-flash,contphy,63.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 10 (p52).
gemini-3.5-flash,crossvid,58.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 10 (p52).
gemini-3.5-flash,livesports_3k,86.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 10 (p52).
gemini-3.5-flash,longvideobench,77.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 10 (p52).
gemini-3.5-flash,lvbench,76.3,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Long/streaming-video chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq50ia3.png).
gemini-3.5-flash,minerva,68.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 10 (p52). Source marks this row with an unexplained dagger. Source marks this row with an unexplained dagger.
gemini-3.5-flash,motionbench,70.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 10 (p52).
gemini-3.5-flash,ovbench,56.5,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Long/streaming-video chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq50ia3.png).
gemini-3.5-flash,ovobench,64.5,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Long/streaming-video chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq50ia3.png).
gemini-3.5-flash,tomato,71.9,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Motion/perception chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4zh7f.png).
gemini-3.5-flash,tvbench,76.4,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Motion/perception chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4zh7f.png).
gemini-3.5-flash,video_mme,87.2,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Long/streaming-video chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq50ia3.png). Source marks this row with an unexplained dagger. Source marks this row with an unexplained dagger.
gemini-3.5-flash,videoholmes,67.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 10 (p52). Source marks this row with an unexplained dagger. Source marks this row with an unexplained dagger.
gemini-3.5-flash,videosimpleqa,76.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 10 (p52).
glm-5.1,code_arena_frontend_elo,1529.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""Arena Code leaderboard"",""judge"":""human preference"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""107,962-vote Arena snapshot"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Code Arena: Frontend snapshot (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4xsvz.jpg). Historical leaderboard snapshot reproduced by ByteDance.
glm-5.2,code_arena_frontend_elo,1593.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""max"",""harness"":""Arena Code leaderboard"",""judge"":""human preference"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""107,962-vote Arena snapshot"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Code Arena: Frontend snapshot (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4xsvz.jpg). Historical leaderboard snapshot reproduced by ByteDance.
gpt-5.5,aethercode,81.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 11 (p53).
gpt-5.5,agent_startup_bench,68.1,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""expert review"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""research and professional-deliverable tools""}",False,official_blog,verified,Release blog EN: General-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4u89c.png).
gpt-5.5,agents_last_exam_average_score,42.8,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""computer-use environment""}",False,official_blog,verified,"Release blog EN: General-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4u89c.png). Turbo is a source dash, not zero."
gpt-5.5,beyond_aime,91.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 11 (p53).
gpt-5.5,blink,78.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
gpt-5.5,charxiv_descriptive,95.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
gpt-5.5,claw_eval_multimodal_pass3,43.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""Pass^3"",""temperature"":""unknown"",""tools"":""high-resolution multimodal agent harness""}",False,official_blog,verified,Release blog EN: Multimodal-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4v1ol.png).
gpt-5.5,creativework,30.5,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""GUI + MCP in Notion, Canva, and Figma""}",False,official_blog,verified,Release blog EN: Computer-use chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4w8x3.png).
gpt-5.5,doubao_multi_turn_bench,62.5,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Daily-life/tool-use chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4ujsp.png).
gpt-5.5,dude,81.7,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
gpt-5.5,dynamath,75.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
gpt-5.5,embspatial_bench,81.9,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Spatial/long-context chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4z2l5.png).
gpt-5.5,emma,79.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
gpt-5.5,erqa,64.5,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Spatial/long-context chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4z2l5.png).
gpt-5.5,frontiercs_algorithmic_v1,40.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""research and code-execution environment""}",False,tech_report,verified,Model card Table 6 (p24). Claude Opus 4.7 is a source dash; official sub-score name unresolved.
gpt-5.5,frontiercs_overall_v1,41.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""research and code-execution environment""}",False,tech_report,verified,Model card Table 6 (p24). Claude Opus 4.7 is a source dash; official sub-score name unresolved.
gpt-5.5,frontiercs_research_v1,58.6,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""research and code-execution environment""}",False,official_blog,verified,Release blog EN: Frontier-research chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq52cdr.png). Claude Opus 4.7 is a source dash; official sub-score name unresolved.
gpt-5.5,fs_researcher,33.9,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 11 (p53).
gpt-5.5,gameworld,34.1,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""browser-game GUI actions""}",False,official_blog,verified,Release blog EN: Computer-use chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4w8x3.png).
gpt-5.5,gdpval_seed_reported_score,84.9,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""source-specific evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""document/spreadsheet deliverable workflow""}",False,official_blog,verified,Release blog EN: General-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4u89c.png).
gpt-5.5,hle_verified,50.4,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Language/search chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq510uc.png).
gpt-5.5,image2floorplan_avg_score,50.7,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""high-resolution multimodal agent harness""}",False,official_blog,verified,Release blog EN: Multimodal-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4v1ol.png).
gpt-5.5,kina,52.6,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Language/search chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq510uc.png).
gpt-5.5,live_mathematician_bench_2026_06,50.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 11 (p53).
gpt-5.5,mathverse_vision_only,84.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
gpt-5.5,mathvista,84.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
gpt-5.5,measurebench,49.9,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Vision capability chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4yer3.png).
gpt-5.5,mmsibench_circular,36.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
gpt-5.5,mobileworld,54.7,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""mobile GUI actions""}",False,official_blog,verified,Release blog EN: Computer-use chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4w8x3.png).
gpt-5.5,msqa,57.4,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Language/search chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq510uc.png).
gpt-5.5,ocrbench_v2,61.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
gpt-5.5,one_million_bench,69.6,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""agentic professional-task environment""}",False,official_blog,verified,Release blog EN: General-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4u89c.png).
gpt-5.5,presentbench,68.9,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""fine-grained rubric evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""slide-generation environment""}",False,official_blog,verified,Release blog EN: General-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4u89c.png).
gpt-5.5,program_bench_almost_resolved_95,5.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official ProgramBench sandbox"",""judge"":""248,000+ behavioral tests"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""compiled executable and documentation""}",False,tech_report,verified,Model card Table 5 (p19).
gpt-5.5,program_bench_fully_resolved,0.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official ProgramBench sandbox"",""judge"":""248,000+ behavioral tests"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""compiled executable and documentation""}",False,tech_report,verified,Model card Table 5 (p19).
gpt-5.5,realworldqa,82.2,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Vision capability chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4yer3.png).
gpt-5.5,seed21_dirtyfilter_precision,94.1,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""300M-token budget; identical validation set"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""rule-only evaluator/optimizer loop""}",False,tech_report,verified,Model card Figure 19 (p32).
gpt-5.5,seed21_dirtyfilter_validation_recall,44.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""300M-token budget; identical validation set"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""rule-only evaluator/optimizer loop""}",False,tech_report,verified,Model card Figure 19 (p32).
gpt-5.5,seedclawbench,66.4,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""OpenClaw-style tools and skills""}",False,official_blog,verified,Release blog EN: Daily-life/tool-use chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4ujsp.png).
gpt-5.5,seedkernelbench_avg_speedup,9.14,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",False,tech_report,verified,Model card Table 5 (p19). Source dashes are not zero.
gpt-5.5,simplevqa,58.6,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
gpt-5.5,supergpqa,72.7,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Language/search chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq510uc.png).
gpt-5.5,swe_atlas,44.7,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",False,official_blog,verified,Release blog EN: Coding chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4wlbt.png).
gpt-5.5,trae_artifacts,56.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",False,tech_report,verified,Model card Table 5 (p19).
gpt-5.5,trae_code_gen_js,76.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",False,tech_report,verified,Model card Table 5 (p19).
gpt-5.5,trae_code_gen_python,73.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",False,tech_report,verified,Model card Table 5 (p19).
gpt-5.5,trae_error_fix_go,70.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",False,tech_report,verified,Model card Table 5 (p19).
gpt-5.5,trae_error_fix_java,65.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",False,tech_report,verified,Model card Table 5 (p19).
gpt-5.5,trae_error_fix_js,79.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",False,tech_report,verified,Model card Table 5 (p19).
gpt-5.5,trae_error_fix_python,76.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",False,tech_report,verified,Model card Table 5 (p19).
gpt-5.5,trae_repo_env,90.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",False,tech_report,verified,Model card Table 5 (p19).
gpt-5.5,treebench,61.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
gpt-5.5,visfactor,56.2,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
gpt-5.5,visulogic,43.5,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
gpt-5.5,vlms_are_biased,49.8,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
gpt-5.5,webbench,81.3,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""Trae or benchmark-specific coding-agent environment""}",False,tech_report,verified,Model card Table 5 (p19).
gpt-5.5,wildclaw_bench_60_openclaw,65.6,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""high-resolution multimodal agent harness""}",False,official_blog,verified,Release blog EN: Multimodal-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4v1ol.png).
gpt-5.5,workspace_bench_openclaw_100,58.7,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""100-task OpenClaw setting"",""judge"":""rubric-based evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""OpenClaw workspace tools""}",False,official_blog,verified,Release blog EN: General-agent chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4u89c.png).
gpt-5.5,workspace_bench_openclaw_100_pass100,20.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""100-task OpenClaw setting"",""judge"":""rubric-based evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""OpenClaw workspace tools""}",False,tech_report,verified,Model card Table 1 (p9).
gpt-5.5,workspace_bench_openclaw_100_pass30,81.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""100-task OpenClaw setting"",""judge"":""rubric-based evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""OpenClaw workspace tools""}",False,tech_report,verified,Model card Table 1 (p9).
gpt-5.5,workspace_bench_openclaw_100_pass50,67.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""100-task OpenClaw setting"",""judge"":""rubric-based evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""OpenClaw workspace tools""}",False,tech_report,verified,Model card Table 1 (p9).
gpt-5.5,workspace_bench_openclaw_100_pass70,44.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""100-task OpenClaw setting"",""judge"":""rubric-based evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""OpenClaw workspace tools""}",False,tech_report,verified,Model card Table 1 (p9).
gpt-5.5,workspace_bench_openclaw_100_pass90,25.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""100-task OpenClaw setting"",""judge"":""rubric-based evaluator"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""OpenClaw workspace tools""}",False,tech_report,verified,Model card Table 1 (p9).
gpt-5.5,worldbench,58.4,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
gpt-5.5,xdailybench,73.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Daily-life/tool-use chart (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4ujsp.png).
gpt-5.5,zerobench_sub,41.0,https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2.1/Seed2_1_Model_Card.pdf,"{""effort"":""unknown"",""harness"":""official"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""unknown"",""tools"":""none""}",False,tech_report,verified,Model card Table 9 (p51).
kimi-k2.6,code_arena_frontend_elo,1513.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""Arena Code leaderboard"",""judge"":""human preference"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""107,962-vote Arena snapshot"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Code Arena: Frontend snapshot (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4xsvz.jpg). Historical leaderboard snapshot reproduced by ByteDance.
minimax-m3,code_arena_frontend_elo,1505.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""unknown"",""harness"":""Arena Code leaderboard"",""judge"":""human preference"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""107,962-vote Arena snapshot"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Code Arena: Frontend snapshot (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4xsvz.jpg). Historical leaderboard snapshot reproduced by ByteDance.
qwen3.7-max,code_arena_frontend_elo,1530.0,https://seed.bytedance.com/en/blog/seed2-1-officially-released-advancing-ai-productivity,"{""effort"":""max"",""harness"":""Arena Code leaderboard"",""judge"":""human preference"",""mode"":""thinking"",""prompt_style"":""official default"",""sampling"":""107,962-vote Arena snapshot"",""temperature"":""unknown"",""tools"":""none""}",False,official_blog,verified,Release blog EN: Code Arena: Frontend snapshot (https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/user-upload/4xfa4mqq4xsvz.jpg). Historical leaderboard snapshot reproduced by ByteDance.
longcat-flash-lite,c_eval,86.55,https://arxiv.org/html/2608.01662v2,"{""context"":""benchmark-specified"",""effort"":""n/a"",""harness"":""official LongCat evaluation; details not disclosed"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official default"",""sampling"":""accuracy"",""temperature"":""official model default"",""tools"":""benchmark-specified""}",True,official_paper,verified,Table 16: C-Eval (acc); accuracy. 
longcat-flash-lite,cmmlu,82.48,https://arxiv.org/html/2608.01662v2,"{""context"":""benchmark-specified"",""effort"":""n/a"",""harness"":""official LongCat evaluation; details not disclosed"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official default"",""sampling"":""accuracy"",""temperature"":""official model default"",""tools"":""benchmark-specified""}",True,official_paper,verified,Table 16: CMMLU (acc); accuracy. 
longcat-flash-lite,gpqa_diamond,66.78,https://arxiv.org/html/2608.01662v2,"{""context"":""benchmark-specified"",""effort"":""n/a"",""harness"":""official LongCat evaluation; details not disclosed"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official default"",""sampling"":""avg@16"",""temperature"":""official model default"",""tools"":""benchmark-specified""}",False,official_paper,verified,Table 16: GPQA-Diamond (avg@16); avg@16. 
longcat-flash-lite,math_500,96.8,https://arxiv.org/html/2608.01662v2,"{""context"":""benchmark-specified"",""effort"":""n/a"",""harness"":""official LongCat evaluation; details not disclosed"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official default"",""sampling"":""accuracy"",""temperature"":""official model default"",""tools"":""benchmark-specified""}",True,official_paper,verified,Table 16: MATH500 (acc); accuracy. 
longcat-flash-lite,mmlu,85.52,https://arxiv.org/html/2608.01662v2,"{""context"":""benchmark-specified"",""effort"":""n/a"",""harness"":""official LongCat evaluation; details not disclosed"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official default"",""sampling"":""accuracy"",""temperature"":""official model default"",""tools"":""benchmark-specified""}",True,official_paper,verified,Table 16: MMLU (acc); accuracy. 
longcat-flash-lite,mmlu_pro,78.29,https://arxiv.org/html/2608.01662v2,"{""context"":""benchmark-specified"",""effort"":""n/a"",""harness"":""official LongCat evaluation; details not disclosed"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official default"",""sampling"":""accuracy"",""temperature"":""official model default"",""tools"":""benchmark-specified""}",True,official_paper,verified,Table 16: MMLU-Pro (acc); accuracy. 
longcat-flash-lite,swe_bench_multilingual,38.1,https://arxiv.org/html/2608.01662v2,"{""context"":""benchmark-specified"",""effort"":""n/a"",""harness"":""official LongCat evaluation; details not disclosed"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official default"",""sampling"":""pass@1; source harness unspecified"",""temperature"":""official model default"",""tools"":""benchmark-specified""}",False,official_paper,verified,Table 16: SWE-Bench Multilingual (acc); pass@1; source harness unspecified. 
longcat-flash-lite,swe_bench_verified,54.4,https://arxiv.org/html/2608.01662v2,"{""context"":""benchmark-specified"",""effort"":""n/a"",""harness"":""official LongCat evaluation; details not disclosed"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official default"",""sampling"":""pass@1; source harness unspecified"",""temperature"":""official model default"",""tools"":""benchmark-specified""}",False,official_paper,verified,Table 16: SWE-Bench Verified (acc); pass@1; source harness unspecified. 
longcat-flash-lite,tau2_bench_telecom,72.8,https://arxiv.org/html/2608.01662v2,"{""context"":""benchmark-specified"",""effort"":""n/a"",""harness"":""official LongCat evaluation; details not disclosed"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official default"",""sampling"":""avg@4"",""temperature"":""official model default"",""tools"":""benchmark-specified""}",False,official_paper,verified,Table 16: τ²-Telecom (avg@4); avg@4. 
longcat-flash-lite,terminal_bench,33.75,https://arxiv.org/html/2608.01662v2,"{""context"":""benchmark-specified"",""effort"":""n/a"",""harness"":""official LongCat evaluation; details not disclosed"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official default"",""sampling"":""Terminal-Bench 2.0 accuracy; harness unspecified"",""temperature"":""official model default"",""tools"":""benchmark-specified""}",False,official_paper,verified,Table 16: TerminalBench 2.0 (acc); Terminal-Bench 2.0 accuracy; harness unspecified. 
longcat-flash-lite,vitabench,7.0,https://arxiv.org/html/2608.01662v2,"{""context"":""benchmark-specified"",""effort"":""n/a"",""harness"":""official LongCat evaluation; details not disclosed"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official default"",""sampling"":""avg@4"",""temperature"":""official model default"",""tools"":""benchmark-specified""}",False,official_paper,verified,Table 16: VitaBench (avg@4); avg@4. 
longcat-flash-lite-sparse,aa_lcr,48.0,https://arxiv.org/html/2608.01662v2,"{""context"":""benchmark-native instance lengths"",""effort"":""n/a"",""harness"":""Official source evaluation at benchmark-native instance lengths"",""hierarchical_indexing"":""off"",""judge"":""deterministic benchmark-native evaluator"",""mode"":""non-thinking"",""prompt_style"":""ATLAS official"",""sampling"":""one response per ATLAS instance"",""temperature"":""official model default"",""tools"":""none""}",False,official_paper,verified,Table 15: AA-LCR; HI=off. 
longcat-flash-lite-sparse,aime_2026,65.73,https://arxiv.org/html/2608.01662v2,"{""context"":""benchmark-specified"",""effort"":""n/a"",""harness"":""official LongCat evaluation; details not disclosed"",""hierarchical_indexing"":""off"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official default"",""sampling"":""avg@32"",""temperature"":""official model default"",""tools"":""benchmark-specified""}",False,official_paper,verified,Table 16: AIME 2026 (avg@32); avg@32. 
longcat-flash-lite-sparse,atlas_amembench_acu_auc_1m,33.25,https://arxiv.org/html/2608.01662v2,"{""context"":""through 1M"",""effort"":""n/a"",""harness"":""ATLAS full eight-slice evaluation through 1M"",""hierarchical_indexing"":""off"",""judge"":""deterministic benchmark-native evaluator"",""mode"":""non-thinking"",""prompt_style"":""ATLAS official"",""sampling"":""one response per ATLAS instance"",""temperature"":""official model default"",""tools"":""none""}",True,official_paper,verified,Table 15: AMemBench-ACU; HI=off. 
longcat-flash-lite-sparse,atlas_graphwalks_extend_auc_1m,66.27,https://arxiv.org/html/2608.01662v2,"{""context"":""through 1M"",""effort"":""n/a"",""harness"":""ATLAS full eight-slice evaluation through 1M"",""hierarchical_indexing"":""off"",""judge"":""deterministic benchmark-native evaluator"",""mode"":""non-thinking"",""prompt_style"":""ATLAS official"",""sampling"":""one response per ATLAS instance"",""temperature"":""official model default"",""tools"":""none""}",True,official_paper,verified,Table 15: GraphWalks Extend; HI=off. 
longcat-flash-lite-sparse,atlas_helmet_icl_extend_auc_1m,91.63,https://arxiv.org/html/2608.01662v2,"{""context"":""through 1M"",""effort"":""n/a"",""harness"":""ATLAS full eight-slice evaluation through 1M"",""hierarchical_indexing"":""off"",""judge"":""deterministic benchmark-native evaluator"",""mode"":""non-thinking"",""prompt_style"":""ATLAS official"",""sampling"":""one response per ATLAS instance"",""temperature"":""official model default"",""tools"":""none""}",True,official_paper,verified,Table 15: HELMET-ICL Extend; HI=off. 
longcat-flash-lite-sparse,atlas_loft_text_retrieval_extend_auc_1m,43.75,https://arxiv.org/html/2608.01662v2,"{""context"":""through 1M"",""effort"":""n/a"",""harness"":""ATLAS full eight-slice evaluation through 1M"",""hierarchical_indexing"":""off"",""judge"":""deterministic benchmark-native evaluator"",""mode"":""non-thinking"",""prompt_style"":""ATLAS official"",""sampling"":""one response per ATLAS instance"",""temperature"":""official model default"",""tools"":""none""}",True,official_paper,verified,Table 15: LOFT Retrieval Extend; HI=off. 
longcat-flash-lite-sparse,atlas_longcodeqa_auc_1m,62.3,https://arxiv.org/html/2608.01662v2,"{""context"":""32K through 1M"",""effort"":""n/a"",""harness"":""ATLAS six-slice evaluation from 32K through 1M"",""hierarchical_indexing"":""off"",""judge"":""deterministic benchmark-native evaluator"",""mode"":""non-thinking"",""prompt_style"":""ATLAS official"",""sampling"":""one response per ATLAS instance"",""temperature"":""official model default"",""tools"":""none""}",True,official_paper,verified,Table 15: LongCodeQA; HI=off. 
longcat-flash-lite-sparse,atlas_mrcr_8needle_auc_1m,44.66,https://arxiv.org/html/2608.01662v2,"{""context"":""through 1M"",""effort"":""n/a"",""harness"":""ATLAS full eight-slice evaluation through 1M"",""hierarchical_indexing"":""off"",""judge"":""deterministic benchmark-native evaluator"",""mode"":""non-thinking"",""prompt_style"":""ATLAS official"",""sampling"":""one response per ATLAS instance"",""temperature"":""official model default"",""tools"":""none""}",True,official_paper,verified,Table 15: MRCR (8-needle); HI=off. 
longcat-flash-lite-sparse,atlas_oolong_synth_auc_1m,38.42,https://arxiv.org/html/2608.01662v2,"{""context"":""through 1M"",""effort"":""n/a"",""harness"":""ATLAS full eight-slice evaluation through 1M"",""hierarchical_indexing"":""off"",""judge"":""deterministic benchmark-native evaluator"",""mode"":""non-thinking"",""prompt_style"":""ATLAS official"",""sampling"":""one response per ATLAS instance"",""temperature"":""official model default"",""tools"":""none""}",True,official_paper,verified,Table 15: OOLong-Synth; HI=off. 
longcat-flash-lite-sparse,beyond_aime,44.2,https://arxiv.org/html/2608.01662v2,"{""context"":""benchmark-specified"",""effort"":""n/a"",""harness"":""official LongCat evaluation; details not disclosed"",""hierarchical_indexing"":""off"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official default"",""sampling"":""avg@10"",""temperature"":""official model default"",""tools"":""benchmark-specified""}",False,official_paper,verified,Table 16: BeyondAIME (avg@10); avg@10. 
longcat-flash-lite-sparse,browsecomp,48.62,https://arxiv.org/html/2608.01662v2,"{""context"":""benchmark-specified"",""effort"":""n/a"",""harness"":""official LongCat evaluation; details not disclosed"",""hierarchical_indexing"":""off"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""official model default"",""tools"":""benchmark-specified""}",True,official_paper,verified,Table 16: BrowseComp (pass@1); pass@1. 
longcat-flash-lite-sparse,browsecomp_zh,61.94,https://arxiv.org/html/2608.01662v2,"{""context"":""benchmark-specified"",""effort"":""n/a"",""harness"":""official LongCat evaluation; details not disclosed"",""hierarchical_indexing"":""off"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official default"",""sampling"":""pass@1 over 289 questions"",""temperature"":""official model default"",""tools"":""benchmark-specified""}",False,official_paper,verified,Table 16: BrowseComp-zh (pass@1); pass@1 over 289 questions. 
longcat-flash-lite-sparse,c_eval,85.76,https://arxiv.org/html/2608.01662v2,"{""context"":""benchmark-specified"",""effort"":""n/a"",""harness"":""official LongCat evaluation; details not disclosed"",""hierarchical_indexing"":""off"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official default"",""sampling"":""accuracy"",""temperature"":""official model default"",""tools"":""benchmark-specified""}",True,official_paper,verified,Table 16: C-Eval (acc); accuracy. 
longcat-flash-lite-sparse,cmmlu,84.25,https://arxiv.org/html/2608.01662v2,"{""context"":""benchmark-specified"",""effort"":""n/a"",""harness"":""official LongCat evaluation; details not disclosed"",""hierarchical_indexing"":""off"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official default"",""sampling"":""accuracy"",""temperature"":""official model default"",""tools"":""benchmark-specified""}",True,official_paper,verified,Table 16: CMMLU (acc); accuracy. 
longcat-flash-lite-sparse,gpqa_diamond,69.49,https://arxiv.org/html/2608.01662v2,"{""context"":""benchmark-specified"",""effort"":""n/a"",""harness"":""official LongCat evaluation; details not disclosed"",""hierarchical_indexing"":""off"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official default"",""sampling"":""avg@16"",""temperature"":""official model default"",""tools"":""benchmark-specified""}",False,official_paper,verified,Table 16: GPQA-Diamond (avg@16); avg@16. 
longcat-flash-lite-sparse,hmmt_feb_2026,40.53,https://arxiv.org/html/2608.01662v2,"{""context"":""benchmark-specified"",""effort"":""n/a"",""harness"":""official LongCat evaluation; details not disclosed"",""hierarchical_indexing"":""off"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official default"",""sampling"":""avg@32"",""temperature"":""official model default"",""tools"":""benchmark-specified""}",False,official_paper,verified,Table 16: HMMT 2026 Feb (avg@32); avg@32. 
longcat-flash-lite-sparse,imo_answerbench,49.38,https://arxiv.org/html/2608.01662v2,"{""context"":""benchmark-specified"",""effort"":""n/a"",""harness"":""official LongCat evaluation; details not disclosed"",""hierarchical_indexing"":""off"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official default"",""sampling"":""avg@4"",""temperature"":""official model default"",""tools"":""benchmark-specified""}",False,official_paper,verified,Table 16: IMO AnswerBench (avg@4); avg@4. 
longcat-flash-lite-sparse,longbench_v2,52.5,https://arxiv.org/html/2608.01662v2,"{""context"":""benchmark-native instance lengths"",""effort"":""n/a"",""harness"":""Official source evaluation at benchmark-native instance lengths"",""hierarchical_indexing"":""off"",""judge"":""deterministic benchmark-native evaluator"",""mode"":""non-thinking"",""prompt_style"":""ATLAS official"",""sampling"":""one response per ATLAS instance"",""temperature"":""official model default"",""tools"":""none""}",False,official_paper,verified,Table 15: LongBench-v2; HI=off. 
longcat-flash-lite-sparse,math_500,95.8,https://arxiv.org/html/2608.01662v2,"{""context"":""benchmark-specified"",""effort"":""n/a"",""harness"":""official LongCat evaluation; details not disclosed"",""hierarchical_indexing"":""off"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official default"",""sampling"":""accuracy"",""temperature"":""official model default"",""tools"":""benchmark-specified""}",True,official_paper,verified,Table 16: MATH500 (acc); accuracy. 
longcat-flash-lite-sparse,mmlu,85.31,https://arxiv.org/html/2608.01662v2,"{""context"":""benchmark-specified"",""effort"":""n/a"",""harness"":""official LongCat evaluation; details not disclosed"",""hierarchical_indexing"":""off"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official default"",""sampling"":""accuracy"",""temperature"":""official model default"",""tools"":""benchmark-specified""}",True,official_paper,verified,Table 16: MMLU (acc); accuracy. 
longcat-flash-lite-sparse,mmlu_pro,79.24,https://arxiv.org/html/2608.01662v2,"{""context"":""benchmark-specified"",""effort"":""n/a"",""harness"":""official LongCat evaluation; details not disclosed"",""hierarchical_indexing"":""off"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official default"",""sampling"":""accuracy"",""temperature"":""official model default"",""tools"":""benchmark-specified""}",True,official_paper,verified,Table 16: MMLU-Pro (acc); accuracy. 
longcat-flash-lite-sparse,rwsearch,68.5,https://arxiv.org/html/2608.01662v2,"{""context"":""benchmark-specified"",""effort"":""n/a"",""harness"":""official LongCat evaluation; details not disclosed"",""hierarchical_indexing"":""off"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official default"",""sampling"":""pass@1"",""temperature"":""official model default"",""tools"":""benchmark-specified""}",True,official_paper,verified,Table 16: RWSearch (pass@1); pass@1. 
longcat-flash-lite-sparse,swe_bench_multilingual,59.33,https://arxiv.org/html/2608.01662v2,"{""context"":""benchmark-specified"",""effort"":""n/a"",""harness"":""official LongCat evaluation; details not disclosed"",""hierarchical_indexing"":""off"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official default"",""sampling"":""pass@1; source harness unspecified"",""temperature"":""official model default"",""tools"":""benchmark-specified""}",False,official_paper,verified,Table 16: SWE-Bench Multilingual (acc); pass@1; source harness unspecified. 
longcat-flash-lite-sparse,swe_bench_pro,40.63,https://arxiv.org/html/2608.01662v2,"{""context"":""benchmark-specified"",""effort"":""n/a"",""harness"":""official LongCat evaluation; details not disclosed"",""hierarchical_indexing"":""off"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official default"",""sampling"":""pass@1; source harness unspecified"",""temperature"":""official model default"",""tools"":""benchmark-specified""}",False,official_paper,verified,Table 16: SWE-Bench Pro (acc); pass@1; source harness unspecified. 
longcat-flash-lite-sparse,swe_bench_verified,68.2,https://arxiv.org/html/2608.01662v2,"{""context"":""benchmark-specified"",""effort"":""n/a"",""harness"":""official LongCat evaluation; details not disclosed"",""hierarchical_indexing"":""off"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official default"",""sampling"":""pass@1; source harness unspecified"",""temperature"":""official model default"",""tools"":""benchmark-specified""}",False,official_paper,verified,Table 16: SWE-Bench Verified (acc); pass@1; source harness unspecified. 
longcat-flash-lite-sparse,tau2_bench_telecom,95.18,https://arxiv.org/html/2608.01662v2,"{""context"":""benchmark-specified"",""effort"":""n/a"",""harness"":""official LongCat evaluation; details not disclosed"",""hierarchical_indexing"":""off"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official default"",""sampling"":""avg@4"",""temperature"":""official model default"",""tools"":""benchmark-specified""}",False,official_paper,verified,Table 16: τ²-Telecom (avg@4); avg@4. 
longcat-flash-lite-sparse,terminal_bench,33.7,https://arxiv.org/html/2608.01662v2,"{""context"":""benchmark-specified"",""effort"":""n/a"",""harness"":""official LongCat evaluation; details not disclosed"",""hierarchical_indexing"":""off"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official default"",""sampling"":""Terminal-Bench 2.0 accuracy; harness unspecified"",""temperature"":""official model default"",""tools"":""benchmark-specified""}",False,official_paper,verified,Table 16: TerminalBench 2.0 (acc); Terminal-Bench 2.0 accuracy; harness unspecified. 
longcat-flash-lite-sparse,vitabench,21.67,https://arxiv.org/html/2608.01662v2,"{""context"":""benchmark-specified"",""effort"":""n/a"",""harness"":""official LongCat evaluation; details not disclosed"",""hierarchical_indexing"":""off"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official default"",""sampling"":""avg@4"",""temperature"":""official model default"",""tools"":""benchmark-specified""}",False,official_paper,verified,Table 16: VitaBench (avg@4); avg@4. 
claude-haiku-4.5,chartqapro,52.87,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 52.87
claude-haiku-4.5,frontier_science_research,17.78,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 17.78
claude-haiku-4.5,frontiersci_olympiad,31.67,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 31.67
claude-haiku-4.5,hle_multimodal,6.85,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 6.85
claude-haiku-4.5,hle_text,8.26,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 8.26
claude-haiku-4.5,ifbench,49.92,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 49.92
claude-haiku-4.5,imo_answerbench,60.75,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 60.75
claude-haiku-4.5,mathvision,65.79,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 65.79
claude-haiku-4.5,microvqa,53.17,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 53.17
claude-haiku-4.5,mmmu_pro,63.12,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 63.12
claude-haiku-4.5,moleculariq,31.74,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 31.74
claude-haiku-4.5,pinchbench,85.77,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 85.77
claude-haiku-4.5,scicode,40.48,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 40.48
claude-haiku-4.5,sfe,48.51,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 48.51
claude-haiku-4.5,sgi_bench,37.11,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 37.11
claude-haiku-4.5,simpleqa_verified,8.0,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Current HF/GitHub image says SimpleQA-Verified; original release image said SimpleQA with visually unchanged values.; Displayed exactly: 8.00
gemini-3.1-flash-lite,chartqapro,67.77,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 67.77
gemini-3.1-flash-lite,frontier_science_research,3.89,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 3.89
gemini-3.1-flash-lite,frontiersci_olympiad,45.41,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 45.41
gemini-3.1-flash-lite,hle_multimodal,14.74,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 14.74
gemini-3.1-flash-lite,hle_text,19.41,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 19.41
gemini-3.1-flash-lite,ifbench,76.15,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 76.15
gemini-3.1-flash-lite,imo_answerbench,64.5,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 64.50
gemini-3.1-flash-lite,mathvision,70.43,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 70.43
gemini-3.1-flash-lite,microvqa,63.34,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 63.34
gemini-3.1-flash-lite,mmlu_pro,87.31,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 87.31
gemini-3.1-flash-lite,mmmu_pro,76.07,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 76.07
gemini-3.1-flash-lite,moleculariq,46.25,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 46.25
gemini-3.1-flash-lite,pinchbench,84.77,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 84.77
gemini-3.1-flash-lite,scicode,41.12,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 41.12
gemini-3.1-flash-lite,sfe,52.03,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 52.03
gemini-3.1-flash-lite,sgi_bench,16.35,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 16.35
gemini-3.1-flash-lite,simpleqa_verified,42.5,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Current HF/GitHub image says SimpleQA-Verified; original release image said SimpleQA with visually unchanged values.; Displayed exactly: 42.50
gemini-3.1-flash-lite,swe_bench_verified,59.6,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""single_value_harness"":""not explicitly identified; likely source-team unified evaluation for cells without slash, but not asserted"",""table_framework_claim"":""OpenCompass/VLMEvalKit for all models conflicts with SWE-specific footnote""}",False,official_model_card,verified,Displayed exactly: 59.60
gemma-4-26b-a4b,chartqapro,44.26,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 44.26
gemma-4-26b-a4b,frontier_science_research,3.33,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 3.33
gemma-4-26b-a4b,frontiersci_olympiad,58.0,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 58.00
gemma-4-26b-a4b,hle_multimodal,9.37,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 9.37
gemma-4-26b-a4b,hle_text,8.7,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 8.70
gemma-4-26b-a4b,imo_answerbench,64.25,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 64.25
gemma-4-26b-a4b,microvqa,63.92,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 63.92
gemma-4-26b-a4b,moleculariq,32.97,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 32.97
gemma-4-26b-a4b,pinchbench,84.13,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 84.13
gemma-4-26b-a4b,sfe,63.09,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 63.09
gemma-4-26b-a4b,sgi_bench,26.1,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 26.10
gemma-4-26b-a4b,simpleqa_verified,9.0,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Current HF/GitHub image says SimpleQA-Verified; original release image said SimpleQA with visually unchanged values.; Displayed exactly: 9.00
gemma-4-26b-a4b,swe_bench_verified,49.8,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""single_value_harness"":""not explicitly identified; likely source-team unified evaluation for cells without slash, but not asserted"",""table_framework_claim"":""OpenCompass/VLMEvalKit for all models conflicts with SWE-specific footnote""}",False,official_model_card,verified,Displayed exactly: 49.80
gpt-5.4-nano,chartqapro,57.5,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 57.50
gpt-5.4-nano,frontier_science_research,5.56,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 5.56
gpt-5.4-nano,frontiersci_olympiad,20.13,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 20.13
gpt-5.4-nano,hle_multimodal,15.15,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 15.15
gpt-5.4-nano,hle_text,19.87,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 19.87
gpt-5.4-nano,ifbench,66.21,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 66.21
gpt-5.4-nano,imo_answerbench,73.5,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 73.50
gpt-5.4-nano,mathvision,71.91,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 71.91
gpt-5.4-nano,microvqa,51.73,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 51.73
gpt-5.4-nano,mmlu_pro,78.91,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 78.91
gpt-5.4-nano,mmmu_pro,63.41,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 63.41
gpt-5.4-nano,moleculariq,45.72,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 45.72
gpt-5.4-nano,pinchbench,79.37,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 79.37
gpt-5.4-nano,scicode,35.5,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 35.50
gpt-5.4-nano,sfe,36.8,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 36.80
gpt-5.4-nano,sgi_bench,25.73,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 25.73
gpt-5.4-nano,simpleqa_verified,11.7,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Current HF/GitHub image says SimpleQA-Verified; original release image said SimpleQA with visually unchanged values.; Displayed exactly: 11.70
gpt-5.4-nano,swe_bench_verified,42.4,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""single_value_harness"":""not explicitly identified; likely source-team unified evaluation for cells without slash, but not asserted"",""table_framework_claim"":""OpenCompass/VLMEvalKit for all models conflicts with SWE-specific footnote""}",False,official_model_card,verified,Displayed exactly: 42.40
intern-s1-pro,chartqapro,64.66,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 64.66
intern-s1-pro,frontier_science_research,12.22,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 12.22
intern-s1-pro,frontiersci_olympiad,53.33,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 53.33
intern-s1-pro,hle_multimodal,10.63,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 10.63
intern-s1-pro,hle_text,19.2,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 19.20
intern-s1-pro,ifbench,69.0,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 69.00
intern-s1-pro,imo_answerbench,77.25,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 77.25
intern-s1-pro,mathvision,72.43,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 72.43
intern-s1-pro,microvqa,63.34,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 63.34
intern-s1-pro,mmlu_pro,86.59,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 86.59
intern-s1-pro,mmmu_pro,72.77,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 72.77
intern-s1-pro,moleculariq,25.31,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 25.31
intern-s1-pro,pinchbench,64.83,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 64.83
intern-s1-pro,scicode,42.74,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 42.74
intern-s1-pro,sfe,52.66,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 52.66
intern-s1-pro,sgi_bench,38.36,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 38.36
intern-s1-pro,simpleqa_verified,41.9,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Current HF/GitHub image says SimpleQA-Verified; original release image said SimpleQA with visually unchanged values.; Displayed exactly: 41.90
intern-s1-pro,swe_bench_verified,38.4,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""single_value_harness"":""not explicitly identified; likely source-team unified evaluation for cells without slash, but not asserted"",""table_framework_claim"":""OpenCompass/VLMEvalKit for all models conflicts with SWE-specific footnote""}",False,official_model_card,verified,Displayed exactly: 38.40
intern-s2-preview-35b,chartqapro,68.9,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""maximum_inference_length_tokens"":64000,""mode"":""thinking enabled by default"",""recommendation_is_exact_eval_setting"":false,""recommended_generation"":{""min_p"":0.0,""temperature"":0.8,""top_k"":50,""top_p"":0.95},""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 68.90
intern-s2-preview-35b,frontier_science_research,19.44,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""maximum_inference_length_tokens"":128000,""mode"":""thinking enabled by default"",""recommendation_is_exact_eval_setting"":false,""recommended_generation"":{""min_p"":0.0,""temperature"":0.8,""top_k"":50,""top_p"":0.95},""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 19.44
intern-s2-preview-35b,frontiersci_olympiad,60.33,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""maximum_inference_length_tokens"":128000,""mode"":""thinking enabled by default"",""recommendation_is_exact_eval_setting"":false,""recommended_generation"":{""min_p"":0.0,""temperature"":0.8,""top_k"":50,""top_p"":0.95},""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 60.33
intern-s2-preview-35b,hle_multimodal,18.07,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""maximum_inference_length_tokens"":64000,""mode"":""thinking enabled by default"",""recommendation_is_exact_eval_setting"":false,""recommended_generation"":{""min_p"":0.0,""temperature"":0.8,""top_k"":50,""top_p"":0.95},""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 18.07
intern-s2-preview-35b,hle_text,21.94,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""maximum_inference_length_tokens"":128000,""mode"":""thinking enabled by default"",""recommendation_is_exact_eval_setting"":false,""recommended_generation"":{""min_p"":0.0,""temperature"":0.8,""top_k"":50,""top_p"":0.95},""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 21.94
intern-s2-preview-35b,ifbench,80.87,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""maximum_inference_length_tokens"":128000,""mode"":""thinking enabled by default"",""recommendation_is_exact_eval_setting"":false,""recommended_generation"":{""min_p"":0.0,""temperature"":0.8,""top_k"":50,""top_p"":0.95},""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 80.87
intern-s2-preview-35b,imo_answerbench,84.0,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""maximum_inference_length_tokens"":128000,""mode"":""thinking enabled by default"",""recommendation_is_exact_eval_setting"":false,""recommended_generation"":{""min_p"":0.0,""temperature"":0.8,""top_k"":50,""top_p"":0.95},""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 84.00
intern-s2-preview-35b,mathvision,83.36,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""maximum_inference_length_tokens"":64000,""mode"":""thinking enabled by default"",""recommendation_is_exact_eval_setting"":false,""recommended_generation"":{""min_p"":0.0,""temperature"":0.8,""top_k"":50,""top_p"":0.95},""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 83.36
intern-s2-preview-35b,microvqa,66.22,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""maximum_inference_length_tokens"":64000,""mode"":""thinking enabled by default"",""recommendation_is_exact_eval_setting"":false,""recommended_generation"":{""min_p"":0.0,""temperature"":0.8,""top_k"":50,""top_p"":0.95},""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 66.22
intern-s2-preview-35b,mmlu_pro,88.0,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""maximum_inference_length_tokens"":128000,""mode"":""thinking enabled by default"",""recommendation_is_exact_eval_setting"":false,""recommended_generation"":{""min_p"":0.0,""temperature"":0.8,""top_k"":50,""top_p"":0.95},""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 88.00
intern-s2-preview-35b,mmmu_pro,76.88,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""maximum_inference_length_tokens"":64000,""mode"":""thinking enabled by default"",""recommendation_is_exact_eval_setting"":false,""recommended_generation"":{""min_p"":0.0,""temperature"":0.8,""top_k"":50,""top_p"":0.95},""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 76.88
intern-s2-preview-35b,moleculariq,57.26,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""maximum_inference_length_tokens"":128000,""mode"":""thinking enabled by default"",""recommendation_is_exact_eval_setting"":false,""recommended_generation"":{""min_p"":0.0,""temperature"":0.8,""top_k"":50,""top_p"":0.95},""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 57.26
intern-s2-preview-35b,pinchbench,88.22,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""maximum_inference_length_tokens"":128000,""mode"":""thinking enabled by default"",""recommendation_is_exact_eval_setting"":false,""recommended_generation"":{""min_p"":0.0,""temperature"":0.8,""top_k"":50,""top_p"":0.95},""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 88.22
intern-s2-preview-35b,scicode,39.64,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""maximum_inference_length_tokens"":128000,""mode"":""thinking enabled by default"",""recommendation_is_exact_eval_setting"":false,""recommended_generation"":{""min_p"":0.0,""temperature"":0.8,""top_k"":50,""top_p"":0.95},""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 39.64
intern-s2-preview-35b,sfe,58.43,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""maximum_inference_length_tokens"":64000,""mode"":""thinking enabled by default"",""recommendation_is_exact_eval_setting"":false,""recommended_generation"":{""min_p"":0.0,""temperature"":0.8,""top_k"":50,""top_p"":0.95},""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 58.43
intern-s2-preview-35b,sgi_bench,52.52,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""maximum_inference_length_tokens"":64000,""mode"":""thinking enabled by default"",""recommendation_is_exact_eval_setting"":false,""recommended_generation"":{""min_p"":0.0,""temperature"":0.8,""top_k"":50,""top_p"":0.95},""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 52.52
intern-s2-preview-35b,simpleqa_verified,27.8,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""maximum_inference_length_tokens"":128000,""mode"":""thinking enabled by default"",""recommendation_is_exact_eval_setting"":false,""recommended_generation"":{""min_p"":0.0,""temperature"":0.8,""top_k"":50,""top_p"":0.95},""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Current HF/GitHub image says SimpleQA-Verified; original release image said SimpleQA with visually unchanged values.; Displayed exactly: 27.80
intern-s2-preview-35b,swe_bench_verified,64.0,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""single_value_harness"":""not explicitly identified; likely source-team unified evaluation for cells without slash, but not asserted"",""table_framework_claim"":""OpenCompass/VLMEvalKit for all models conflicts with SWE-specific footnote""}",False,official_model_card,verified,Displayed exactly: 64.00
qwen3.6-35b-a3b,chartqapro,64.54,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 64.54
qwen3.6-35b-a3b,frontier_science_research,10.0,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 10.00
qwen3.6-35b-a3b,frontiersci_olympiad,65.0,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 65.00
qwen3.6-35b-a3b,hle_multimodal,17.04,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 17.04
qwen3.6-35b-a3b,hle_text,21.42,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 21.42
qwen3.6-35b-a3b,ifbench,63.52,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 63.52
qwen3.6-35b-a3b,imo_answerbench,81.0,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 81.00
qwen3.6-35b-a3b,mathvision,80.59,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 80.59
qwen3.6-35b-a3b,microvqa,63.63,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 63.63
qwen3.6-35b-a3b,mmlu_pro,85.12,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 85.12
qwen3.6-35b-a3b,mmmu_pro,74.28,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 74.28
qwen3.6-35b-a3b,moleculariq,32.62,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 32.62
qwen3.6-35b-a3b,pinchbench,87.05,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 87.05
qwen3.6-35b-a3b,scicode,40.6,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 40.60
qwen3.6-35b-a3b,sfe,51.31,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 51.31
qwen3.6-35b-a3b,sgi_bench,37.3,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 37.30
qwen3.6-35b-a3b,simpleqa_verified,20.8,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Current HF/GitHub image says SimpleQA-Verified; original release image said SimpleQA with visually unchanged values.; Displayed exactly: 20.80
qwen3.6-35b-a3b,swe_bench_verified,73.4,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""after_slash"":""AgentCompass unified framework with Mini-SWE-Agent"",""before_slash"":""officially reported in model technical report"",""footnote"":""[2]""}",False,official_model_card,verified,"Displayed exactly: 73.40 / 62.80[2]; Alternatives: [{""footnote"": ""[2]"", ""role"": ""official_report"", ""value"": ""73.40""}, {""footnote"": ""[2]"", ""harness"": ""AgentCompass with Mini-SWE-Agent"", ""role"": ""unified_re_evaluation"", ""value"": ""62.80""}]"
step-3.5-flash,frontier_science_research,10.0,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 10.00
step-3.5-flash,frontiersci_olympiad,70.0,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 70.00
step-3.5-flash,hle_text,23.04,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 23.04
step-3.5-flash,ifbench,65.72,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 65.72
step-3.5-flash,imo_answerbench,79.0,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""harness"":""OpenCompass for displayed 79.00 per footnote [1]"",""official_report_sampling"":""pass@1 averaged over 8 independent generations/problem"",""official_report_value"":""85.4"",""source"":""arXiv:2602.10604v2""}",False,official_model_card,verified,"Displayed exactly: 79.00[1]; Alternatives: [{""footnote"": ""[1]"", ""role"": ""displayed_OpenCompass_evaluation"", ""value"": ""79.00""}, {""footnote"": ""[1]"", ""role"": ""official_technical_report"", ""sampling"": ""average of 8 generations/problem"", ""source_url"": ""https://arxiv.org/pdf/2602.10604v2"", ""value"": ""85.4""}]"
step-3.5-flash,mmlu_pro,83.44,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 83.44
step-3.5-flash,moleculariq,45.94,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 45.94
step-3.5-flash,pinchbench,85.0,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 85.00
step-3.5-flash,scicode,46.15,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 46.15
step-3.5-flash,sgi_bench,36.16,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Displayed exactly: 36.16
step-3.5-flash,simpleqa_verified,22.1,https://huggingface.co/internlm/Intern-S2-Preview/resolve/3b53ea8d2f8ed46abefa507e3f6f56ee0c662c4a/figs/performance.png,"{""exact_per_cell_settings"":""not stated"",""table_framework_claim"":""OpenCompass and VLMEvalKit used to evaluate all models""}",False,official_model_card,verified,Current HF/GitHub image says SimpleQA-Verified; original release image said SimpleQA with visually unchanged values.; Displayed exactly: 22.10
step-3.5-flash,swe_bench_verified,74.4,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""% resolved"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""source label only"",""effort"":""source does not state"",""harness_agent"":""source does not state coding scaffold"",""judge"":""benchmark-specified; exact StepFun judge not stated"",""mode"":""source does not state"",""multimodal_input"":false,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""agentic coding tools"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",True,official_blog,verified,Official StepFun launch comparison table.
claude-opus-4.5,lmarena_search_elo,1181.0,https://ernie.baidu.com/blog/posts/ernie-5.1-0508-release/lmarena-search-ranking.png,"{""leaderboard"":""Arena AI Search"",""metric"":""Arena score"",""snapshot_date"":""2026-05-09"",""style_control"":""off"",""tools"":""search arena environment""}",False,official_blog,verified,"May 9, 2026 Search Arena snapshot; style control off. Research observation obs-047. Displayed rank 15. Baidu chart footer cites ARENA.AI/LEADERBOARD/SEARCH."
claude-opus-4.6,lmarena_search_elo,1255.0,https://ernie.baidu.com/blog/posts/ernie-5.1-0508-release/lmarena-search-ranking.png,"{""leaderboard"":""Arena AI Search"",""metric"":""Arena score"",""snapshot_date"":""2026-05-09"",""style_control"":""off"",""tools"":""search arena environment""}",False,official_blog,verified,"May 9, 2026 Search Arena snapshot; style control off. Research observation obs-033. Displayed rank 1. Baidu chart footer cites ARENA.AI/LEADERBOARD/SEARCH."
claude-opus-4.6,spreadsheetbench_verified,82.4,https://ernie.baidu.com/blog/posts/ernie-5.1-0508-release/benchmark.png,"{""sampling"":""pass@1"",""tools"":""code execution/spreadsheet agent""}",False,official_blog,verified,"Verified split, pass@1, with spreadsheet/code execution. Research observation obs-015."
claude-opus-4.7,chatbot_arena_elo,1503.0,https://ernie.baidu.com/blog/posts/ernie-5.1-0508-release/creative-capability.png,"{""leaderboard"":""Arena AI Text"",""metric"":""Arena score"",""mode"":""thinking"",""snapshot_date"":""2026-04-30"",""style_control"":""on""}",False,official_blog,verified,"April 30, 2026 Text Arena snapshot; style control on. Research observation obs-048. Displayed rank 1. Identical asset is published in four page placements. The same PNG bytes are published under two filenames and four English/Chinese page placements."
claude-opus-4.7,lmarena_search_elo,1236.0,https://ernie.baidu.com/blog/posts/ernie-5.1-0508-release/lmarena-search-ranking.png,"{""leaderboard"":""Arena AI Search"",""metric"":""Arena score"",""mode"":""unspecified"",""snapshot_date"":""2026-05-09"",""style_control"":""off"",""tools"":""search arena environment""}",False,official_blog,verified,"May 9, 2026 Search Arena snapshot; style control off. Research observation obs-035. Displayed rank 3. Baidu chart footer cites ARENA.AI/LEADERBOARD/SEARCH."
claude-sonnet-4.6,lmarena_search_elo,1221.0,https://ernie.baidu.com/blog/posts/ernie-5.1-0508-release/lmarena-search-ranking.png,"{""leaderboard"":""Arena AI Search"",""metric"":""Arena score"",""snapshot_date"":""2026-05-09"",""style_control"":""off"",""tools"":""search arena environment""}",False,official_blog,verified,"May 9, 2026 Search Arena snapshot; style control off. Research observation obs-037. Displayed rank 5. Baidu chart footer cites ARENA.AI/LEADERBOARD/SEARCH."
deepseek-v4-pro,spreadsheetbench_verified,67.0,https://ernie.baidu.com/blog/posts/ernie-5.1-0508-release/benchmark.png,"{""sampling"":""pass@1"",""tools"":""code execution/spreadsheet agent""}",False,official_blog,verified,"Verified split, pass@1, with spreadsheet/code execution. Research observation obs-014."
ernie-5.1,aime_2026,99.6,https://ernie.baidu.com/blog/posts/ernie-5.1-0508-release/benchmark.png,"{""metric"":""score"",""sampling"":""avg@32"",""tools"":""tools enabled (unspecified)""}",False,official_blog,verified,With-tool avg@32 value is noncanonical against pass@1/no-tools. Research observation obs-001.
ernie-5.1,deepsearchqa_f1,77.3,https://ernie.baidu.com/blog/posts/ernie-5.1-0508-release/benchmark.png,"{""metric"":""F1"",""tools"":""search/agentic""}",False,official_blog,verified,F1 search-agent row matches the campaign benchmark identity. Research observation obs-009.
ernie-5.1,lmarena_search_elo,1223.0,https://ernie.baidu.com/blog/posts/ernie-5.1-0508-release/lmarena-search-ranking.png,"{""leaderboard"":""Arena AI Search"",""metric"":""Arena score"",""snapshot_date"":""2026-05-09"",""style_control"":""off"",""tools"":""search arena environment""}",False,official_blog,verified,"May 9, 2026 Search Arena snapshot; style control off. Research observation obs-036. Displayed rank 4. Baidu chart footer cites ARENA.AI/LEADERBOARD/SEARCH."
ernie-5.1,mmlu_pro,84.3,https://ernie.baidu.com/blog/posts/ernie-5.1-0508-release/benchmark.png,"{""metric"":""accuracy""}",False,official_blog,verified,Accuracy row matches MMLU-Pro; source does not disclose tools. Research observation obs-025.
ernie-5.1,spreadsheetbench_verified,72.5,https://ernie.baidu.com/blog/posts/ernie-5.1-0508-release/benchmark.png,"{""sampling"":""pass@1"",""tools"":""code execution/spreadsheet agent""}",False,official_blog,verified,"Verified split, pass@1, with spreadsheet/code execution. Research observation obs-013."
ernie-5.1,tau3_bench,67.9,https://ernie.baidu.com/blog/posts/ernie-5.1-0508-release/benchmark.png,"{""sampling"":""avg@4"",""tools"":""agentic""}",False,official_blog,verified,The source's avg@4 agentic setting matches the campaign row. Research observation obs-005.
ernie-5.1-preview,chatbot_arena_elo,1476.0,https://ernie.baidu.com/blog/posts/ernie-5.1-0508-release/creative-capability.png,"{""leaderboard"":""Arena AI Text"",""metric"":""Arena score"",""snapshot_date"":""2026-04-30"",""style_control"":""on""}",False,official_blog,verified,"April 30, 2026 Text Arena snapshot; style control on. Research observation obs-060. Displayed rank 13. Identical asset is published in four page placements. The same PNG bytes are published under two filenames and four English/Chinese page placements."
gemini-3-flash,lmarena_search_elo,1209.0,https://ernie.baidu.com/blog/posts/ernie-5.1-0508-release/lmarena-search-ranking.png,"{""leaderboard"":""Arena AI Search"",""metric"":""Arena score"",""snapshot_date"":""2026-05-09"",""style_control"":""off"",""tools"":""grounding/search""}",False,official_blog,verified,"May 9, 2026 Search Arena snapshot; style control off. Research observation obs-041. Displayed rank 9. Baidu chart footer cites ARENA.AI/LEADERBOARD/SEARCH."
gemini-3-pro,lmarena_search_elo,1209.0,https://ernie.baidu.com/blog/posts/ernie-5.1-0508-release/lmarena-search-ranking.png,"{""leaderboard"":""Arena AI Search"",""metric"":""Arena score"",""snapshot_date"":""2026-05-09"",""style_control"":""off"",""tools"":""grounding/search""}",False,official_blog,verified,"May 9, 2026 Search Arena snapshot; style control off. Research observation obs-040. Displayed rank 8. Baidu chart footer cites ARENA.AI/LEADERBOARD/SEARCH."
gemini-3.1-pro,lmarena_search_elo,1217.0,https://ernie.baidu.com/blog/posts/ernie-5.1-0508-release/lmarena-search-ranking.png,"{""leaderboard"":""Arena AI Search"",""metric"":""Arena score"",""snapshot_date"":""2026-05-09"",""style_control"":""off"",""tools"":""grounding/search""}",False,official_blog,verified,"May 9, 2026 Search Arena snapshot; style control off. Research observation obs-038. Displayed rank 6. Baidu chart footer cites ARENA.AI/LEADERBOARD/SEARCH."
gemini-3.1-pro,spreadsheetbench_verified,80.8,https://ernie.baidu.com/blog/posts/ernie-5.1-0508-release/benchmark.png,"{""sampling"":""pass@1"",""tools"":""code execution/spreadsheet agent""}",False,official_blog,verified,"Verified split, pass@1, with spreadsheet/code execution. Research observation obs-016."
gpt-5.1,lmarena_search_elo,1200.0,https://ernie.baidu.com/blog/posts/ernie-5.1-0508-release/lmarena-search-ranking.png,"{""leaderboard"":""Arena AI Search"",""metric"":""Arena score"",""snapshot_date"":""2026-05-09"",""style_control"":""off"",""tools"":""search arena environment""}",False,official_blog,verified,"May 9, 2026 Search Arena snapshot; style control off. Research observation obs-044. Displayed rank 12. Baidu chart footer cites ARENA.AI/LEADERBOARD/SEARCH."
gpt-5.2,lmarena_search_elo,1213.0,https://ernie.baidu.com/blog/posts/ernie-5.1-0508-release/lmarena-search-ranking.png,"{""leaderboard"":""Arena AI Search"",""metric"":""Arena score"",""snapshot_date"":""2026-05-09"",""style_control"":""off"",""tools"":""search arena environment""}",False,official_blog,verified,"May 9, 2026 Search Arena snapshot; style control off. Research observation obs-039. Displayed rank 7. Baidu chart footer cites ARENA.AI/LEADERBOARD/SEARCH."
gpt-5.4,chatbot_arena_elo,1478.0,https://ernie.baidu.com/blog/posts/ernie-5.1-0508-release/creative-capability.png,"{""effort"":""high"",""leaderboard"":""Arena AI Text"",""metric"":""Arena score"",""snapshot_date"":""2026-04-30"",""style_control"":""on""}",False,official_blog,verified,"April 30, 2026 Text Arena snapshot; style control on. Research observation obs-057. Displayed rank 10. Identical asset is published in four page placements. The same PNG bytes are published under two filenames and four English/Chinese page placements."
gpt-5.4,lmarena_search_elo,1201.0,https://ernie.baidu.com/blog/posts/ernie-5.1-0508-release/lmarena-search-ranking.png,"{""leaderboard"":""Arena AI Search"",""metric"":""Arena score"",""snapshot_date"":""2026-05-09"",""style_control"":""off"",""tools"":""search arena environment""}",False,official_blog,verified,"May 9, 2026 Search Arena snapshot; style control off. Research observation obs-043. Displayed rank 11. Baidu chart footer cites ARENA.AI/LEADERBOARD/SEARCH."
gpt-5.5,chatbot_arena_elo,1488.0,https://ernie.baidu.com/blog/posts/ernie-5.1-0508-release/creative-capability.png,"{""effort"":""high"",""leaderboard"":""Arena AI Text"",""metric"":""Arena score"",""snapshot_date"":""2026-04-30"",""style_control"":""on""}",False,official_blog,verified,"April 30, 2026 Text Arena snapshot; style control on. Research observation obs-054. Displayed rank 7. Identical asset is published in four page placements. The same PNG bytes are published under two filenames and four English/Chinese page placements."
gpt-5.5,lmarena_search_elo,1242.0,https://ernie.baidu.com/blog/posts/ernie-5.1-0508-release/lmarena-search-ranking.png,"{""leaderboard"":""Arena AI Search"",""metric"":""Arena score"",""snapshot_date"":""2026-05-09"",""style_control"":""off"",""tools"":""search arena environment""}",False,official_blog,verified,"May 9, 2026 Search Arena snapshot; style control off. Research observation obs-034. Displayed rank 2. Baidu chart footer cites ARENA.AI/LEADERBOARD/SEARCH."
grok-4.20,chatbot_arena_elo,1479.0,https://ernie.baidu.com/blog/posts/ernie-5.1-0508-release/creative-capability.png,"{""leaderboard"":""Arena AI Text"",""metric"":""Arena score"",""mode"":""unspecified"",""snapshot_date"":""2026-04-30"",""style_control"":""on""}",False,official_blog,verified,"April 30, 2026 Text Arena snapshot; style control on. Research observation obs-056. Displayed rank 9. Identical asset is published in four page placements. The same PNG bytes are published under two filenames and four English/Chinese page placements."
grok-4.20,lmarena_search_elo,1209.0,https://ernie.baidu.com/blog/posts/ernie-5.1-0508-release/lmarena-search-ranking.png,"{""leaderboard"":""Arena AI Search"",""metric"":""Arena score"",""mode"":""multi-agent"",""snapshot_date"":""2026-05-09"",""style_control"":""off"",""tools"":""search arena environment""}",False,official_blog,verified,"May 9, 2026 Search Arena snapshot; style control off. Research observation obs-042. Displayed rank 10. Baidu chart footer cites ARENA.AI/LEADERBOARD/SEARCH."
grok-4.3,lmarena_search_elo,1200.0,https://ernie.baidu.com/blog/posts/ernie-5.1-0508-release/lmarena-search-ranking.png,"{""leaderboard"":""Arena AI Search"",""metric"":""Arena score"",""snapshot_date"":""2026-05-09"",""style_control"":""off"",""tools"":""search arena environment""}",False,official_blog,verified,"May 9, 2026 Search Arena snapshot; style control off. Research observation obs-045. Displayed rank 13. Baidu chart footer cites ARENA.AI/LEADERBOARD/SEARCH."
muse-spark,chatbot_arena_elo,1489.0,https://ernie.baidu.com/blog/posts/ernie-5.1-0508-release/creative-capability.png,"{""leaderboard"":""Arena AI Text"",""metric"":""Arena score"",""snapshot_date"":""2026-04-30"",""style_control"":""on""}",False,official_blog,verified,"April 30, 2026 Text Arena snapshot; style control on. Research observation obs-053. Displayed rank 6. Identical asset is published in four page placements. The same PNG bytes are published under two filenames and four English/Chinese page placements."
claude-haiku-4.5,aa_lcr_mistral_custom_gpt_4_1_mini_middle_out,80.0,https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/resolve/a11f36bebf709121056b1dbcc943d1c6afbe494d/images/aime.png?download=true,"{""context"":""source does not state"",""effort"":""128K thinking budget"",""harness"":""Mistral custom AA-LCR with middle-out context handling"",""judge"":""gpt-4.1-mini-2025-04-14 equality judge"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1; repeat count undisclosed by Mistral"",""temperature"":""default (Anthropic default)"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly: '80'. Research observation: small-aime.png:1:3:score:reasoning. Chart footnotes: *Custom implementation of AA LCR using gpt-4.1-mini-2025-04-14 as a judge and a middle-out approach for models with shorter context lengths.
claude-sonnet-4.5,tau3_airline,72.0,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image4.png?download=true,"{""context"":""200K (default)"",""effort"":""high (maximum)"",""harness"":""Sierra-reported \u03c4\u00b3 result"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""interleaved scratchpads"",""sampling"":""Sierra-reported; trial count not stated here"",""temperature"":""default (top_p, temperature)"",""tools"":""\u03c4\u00b3 domain tools and retrieval environment""}",False,official_model_card,verified,"Displayed exactly: '72.0'. Research observation: medium-image4.png:3:2:score:reported. Chart footnotes: *All model evaluations were run with maximum reasoning settings. | **Sonnet 4.6 encountered higher rates of reasoning truncation due to external API restrictions than other models, which affected its performance. | *** Self-reported | **** Self-reported, Mistral is using context management and a discard-all strategy at 100k tokens. | τ³ scores as reported by Sierra for Claude Sonnet 4.5 and Qwen3.5. Others with user simulator: gpt-5.2 with reasoning_effort: low. 4 trials. Banking domain evaluated with terminal- or embedding-based agentic search retrieval, only highest score is reported."
claude-sonnet-4.5,tau3_banking,22.4,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image4.png?download=true,"{""context"":""200K (default)"",""effort"":""high (maximum)"",""harness"":""Sierra-reported \u03c4\u00b3 result"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""interleaved scratchpads"",""sampling"":""Sierra-reported; trial count not stated here"",""temperature"":""default (top_p, temperature)"",""tools"":""\u03c4\u00b3 Banking terminal or embedding retrieval; higher of the two reported""}",False,official_model_card,verified,"Displayed exactly: '22.4'. Research observation: medium-image4.png:5:2:score:reported. Chart footnotes: *All model evaluations were run with maximum reasoning settings. | **Sonnet 4.6 encountered higher rates of reasoning truncation due to external API restrictions than other models, which affected its performance. | *** Self-reported | **** Self-reported, Mistral is using context management and a discard-all strategy at 100k tokens. | τ³ scores as reported by Sierra for Claude Sonnet 4.5 and Qwen3.5. Others with user simulator: gpt-5.2 with reasoning_effort: low. 4 trials. Banking domain evaluated with terminal- or embedding-based agentic search retrieval, only highest score is reported."
claude-sonnet-4.5,tau3_retail,72.4,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image4.png?download=true,"{""context"":""200K (default)"",""effort"":""high (maximum)"",""harness"":""Sierra-reported \u03c4\u00b3 result"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""interleaved scratchpads"",""sampling"":""Sierra-reported; trial count not stated here"",""temperature"":""default (top_p, temperature)"",""tools"":""\u03c4\u00b3 domain tools and retrieval environment""}",False,official_model_card,verified,"Displayed exactly: '72.4'. Research observation: medium-image4.png:4:2:score:reported. Chart footnotes: *All model evaluations were run with maximum reasoning settings. | **Sonnet 4.6 encountered higher rates of reasoning truncation due to external API restrictions than other models, which affected its performance. | *** Self-reported | **** Self-reported, Mistral is using context management and a discard-all strategy at 100k tokens. | τ³ scores as reported by Sierra for Claude Sonnet 4.5 and Qwen3.5. Others with user simulator: gpt-5.2 with reasoning_effort: low. 4 trials. Banking domain evaluated with terminal- or embedding-based agentic search retrieval, only highest score is reported."
claude-sonnet-4.5,tau3_telecom,84.9,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image4.png?download=true,"{""context"":""200K (default)"",""effort"":""high (maximum)"",""harness"":""Sierra-reported \u03c4\u00b3 result"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""interleaved scratchpads"",""sampling"":""Sierra-reported; trial count not stated here"",""temperature"":""default (top_p, temperature)"",""tools"":""\u03c4\u00b3 domain tools and retrieval environment""}",False,official_model_card,verified,"Displayed exactly: '84.9'. Research observation: medium-image4.png:2:2:score:reported. Chart footnotes: *All model evaluations were run with maximum reasoning settings. | **Sonnet 4.6 encountered higher rates of reasoning truncation due to external API restrictions than other models, which affected its performance. | *** Self-reported | **** Self-reported, Mistral is using context management and a discard-all strategy at 100k tokens. | τ³ scores as reported by Sierra for Claude Sonnet 4.5 and Qwen3.5. Others with user simulator: gpt-5.2 with reasoning_effort: low. 4 trials. Banking domain evaluated with terminal- or embedding-based agentic search retrieval, only highest score is reported."
claude-sonnet-4.6,beyond_aime,47.3,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image1.png?download=true,"{""context"":""default"",""effort"":""max"",""harness"":""Mistral official maximum-reasoning comparison"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""samples=16 (reported as avg@16)"",""temperature"":""default"",""tools"":""none""}",False,official_model_card,verified,"Displayed exactly: '47.3'. Research observation: medium-image1.png:4:3:score:reported. Chart footnotes: *All model evaluations were run with maximum reasoning settings. | **Sonnet 4.6 encountered higher rates of reasoning truncation due to external API restrictions than other models, which affected its performance."
claude-sonnet-4.6,collie,67.7,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image1.png?download=true,"{""context"":""default"",""effort"":""max"",""harness"":""Mistral official maximum-reasoning comparison"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1; repeat count undisclosed"",""temperature"":""default"",""tools"":""none""}",False,official_model_card,verified,"Displayed exactly: '67.7'. Research observation: medium-image1.png:3:3:score:reported. Chart footnotes: *All model evaluations were run with maximum reasoning settings. | **Sonnet 4.6 encountered higher rates of reasoning truncation due to external API restrictions than other models, which affected its performance."
claude-sonnet-4.6,ifbench,57.1,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image1.png?download=true,"{""context"":""default"",""effort"":""max"",""harness"":""Mistral official maximum-reasoning comparison"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1; repeat count undisclosed"",""temperature"":""default"",""tools"":""none""}",False,official_model_card,verified,"Displayed exactly: '57.1'. Research observation: medium-image1.png:2:3:score:reported. Chart footnotes: *All model evaluations were run with maximum reasoning settings. | **Sonnet 4.6 encountered higher rates of reasoning truncation due to external API restrictions than other models, which affected its performance."
claude-sonnet-4.6,tau3_airline,83.0,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image4.png?download=true,"{""context"":""default"",""effort"":""max"",""harness"":""Mistral \u03c4\u00b3 evaluation with GPT-5.2 low-effort user simulator"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""trials=4"",""temperature"":""default"",""tools"":""\u03c4\u00b3 domain tools and retrieval environment""}",False,official_model_card,verified,"Displayed exactly: '83.0'. Research observation: medium-image4.png:3:3:score:reported. Chart footnotes: *All model evaluations were run with maximum reasoning settings. | **Sonnet 4.6 encountered higher rates of reasoning truncation due to external API restrictions than other models, which affected its performance. | *** Self-reported | **** Self-reported, Mistral is using context management and a discard-all strategy at 100k tokens. | τ³ scores as reported by Sierra for Claude Sonnet 4.5 and Qwen3.5. Others with user simulator: gpt-5.2 with reasoning_effort: low. 4 trials. Banking domain evaluated with terminal- or embedding-based agentic search retrieval, only highest score is reported."
claude-sonnet-4.6,tau3_banking,28.4,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image4.png?download=true,"{""context"":""default"",""effort"":""max"",""harness"":""Mistral \u03c4\u00b3 evaluation with GPT-5.2 low-effort user simulator"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""trials=4"",""temperature"":""default"",""tools"":""\u03c4\u00b3 Banking terminal or embedding retrieval; higher of the two reported""}",False,official_model_card,verified,"Displayed exactly: '28.4'. Research observation: medium-image4.png:5:3:score:reported. Chart footnotes: *All model evaluations were run with maximum reasoning settings. | **Sonnet 4.6 encountered higher rates of reasoning truncation due to external API restrictions than other models, which affected its performance. | *** Self-reported | **** Self-reported, Mistral is using context management and a discard-all strategy at 100k tokens. | τ³ scores as reported by Sierra for Claude Sonnet 4.5 and Qwen3.5. Others with user simulator: gpt-5.2 with reasoning_effort: low. 4 trials. Banking domain evaluated with terminal- or embedding-based agentic search retrieval, only highest score is reported."
claude-sonnet-4.6,tau3_retail,75.9,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image4.png?download=true,"{""context"":""default"",""effort"":""max"",""harness"":""Mistral \u03c4\u00b3 evaluation with GPT-5.2 low-effort user simulator"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""trials=4"",""temperature"":""default"",""tools"":""\u03c4\u00b3 domain tools and retrieval environment""}",False,official_model_card,verified,"Displayed exactly: '75.9'. Research observation: medium-image4.png:4:3:score:reported. Chart footnotes: *All model evaluations were run with maximum reasoning settings. | **Sonnet 4.6 encountered higher rates of reasoning truncation due to external API restrictions than other models, which affected its performance. | *** Self-reported | **** Self-reported, Mistral is using context management and a discard-all strategy at 100k tokens. | τ³ scores as reported by Sierra for Claude Sonnet 4.5 and Qwen3.5. Others with user simulator: gpt-5.2 with reasoning_effort: low. 4 trials. Banking domain evaluated with terminal- or embedding-based agentic search retrieval, only highest score is reported."
claude-sonnet-4.6,tau3_telecom,70.4,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image4.png?download=true,"{""context"":""default"",""effort"":""max"",""harness"":""Mistral \u03c4\u00b3 evaluation with GPT-5.2 low-effort user simulator"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""trials=4"",""temperature"":""default"",""tools"":""\u03c4\u00b3 domain tools and retrieval environment""}",False,official_model_card,verified,"Displayed exactly: '70.4'. Research observation: medium-image4.png:2:3:score:reported. Chart footnotes: *All model evaluations were run with maximum reasoning settings. | **Sonnet 4.6 encountered higher rates of reasoning truncation due to external API restrictions than other models, which affected its performance. | *** Self-reported | **** Self-reported, Mistral is using context management and a discard-all strategy at 100k tokens. | τ³ scores as reported by Sierra for Claude Sonnet 4.5 and Qwen3.5. Others with user simulator: gpt-5.2 with reasoning_effort: low. 4 trials. Banking domain evaluated with terminal- or embedding-based agentic search retrieval, only highest score is reported."
devstral-small-2,swe_bench_verified,68.0,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image3.png?download=true,"{""context"":""256k"",""effort"":""none"",""harness"":""Mistral self-reported SWE-bench Verified evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official CHAT_SYSTEM_PROMPT.txt"",""sampling"":""source does not state trial count"",""temperature"":""0.2 release recommendation; immutable HF examples use 0.15"",""tools"":""agentic coding scaffold and SWE-bench verifier""}",False,official_model_card,verified,Displayed exactly: '68.0'. Research observation: medium-image3.png:1:3:score:reported.
glm-5,aime_2025,87.1,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image1.png?download=true,"{""context"":""default"",""effort"":""default (maximum reasoning setting)"",""harness"":""Mistral official maximum-reasoning comparison"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""samples=16 (reported as avg@16)"",""temperature"":""default"",""tools"":""none""}",False,official_model_card,verified,"Displayed exactly: '87.1'. Research observation: medium-image1.png:1:6:score:reported. Chart footnotes: *All model evaluations were run with maximum reasoning settings. | **Sonnet 4.6 encountered higher rates of reasoning truncation due to external API restrictions than other models, which affected its performance."
glm-5,collie,86.4,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image1.png?download=true,"{""context"":""default"",""effort"":""default (maximum reasoning setting)"",""harness"":""Mistral official maximum-reasoning comparison"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1; repeat count undisclosed"",""temperature"":""default"",""tools"":""none""}",False,official_model_card,verified,"Displayed exactly: '86.4'. Research observation: medium-image1.png:3:6:score:reported. Chart footnotes: *All model evaluations were run with maximum reasoning settings. | **Sonnet 4.6 encountered higher rates of reasoning truncation due to external API restrictions than other models, which affected its performance."
glm-5,ifbench,67.0,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image1.png?download=true,"{""context"":""default"",""effort"":""default (maximum reasoning setting)"",""harness"":""Mistral official maximum-reasoning comparison"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1; repeat count undisclosed"",""temperature"":""default"",""tools"":""none""}",False,official_model_card,verified,"Displayed exactly: '67.0'. Research observation: medium-image1.png:2:6:score:reported. Chart footnotes: *All model evaluations were run with maximum reasoning settings. | **Sonnet 4.6 encountered higher rates of reasoning truncation due to external API restrictions than other models, which affected its performance."
gpt-oss-120b,aa_lcr_mistral_custom_gpt_4_1_mini_middle_out,51.0,https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/resolve/a11f36bebf709121056b1dbcc943d1c6afbe494d/images/aime.png?download=true,"{""context"":""default (128K)"",""effort"":""chart reasoning mode; exact effort level undisclosed"",""harness"":""Mistral custom AA-LCR with middle-out context handling"",""judge"":""gpt-4.1-mini-2025-04-14 equality judge"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1; repeat count undisclosed by Mistral"",""temperature"":""default"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly: '51'. Research observation: small-aime.png:1:2:score:reasoning. Chart footnotes: *Custom implementation of AA LCR using gpt-4.1-mini-2025-04-14 as a judge and a middle-out approach for models with shorter context lengths.
kimi-k2.5,beyond_aime,60.3,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image1.png?download=true,"{""context"":""256K"",""effort"":""default (maximum reasoning setting)"",""harness"":""Mistral official maximum-reasoning comparison"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""samples=16 (reported as avg@16)"",""temperature"":""1.0"",""tools"":""none""}",False,official_model_card,verified,"Displayed exactly: '60.3'. Research observation: medium-image1.png:4:4:score:reported. Chart footnotes: *All model evaluations were run with maximum reasoning settings. | **Sonnet 4.6 encountered higher rates of reasoning truncation due to external API restrictions than other models, which affected its performance."
kimi-k2.5,collie,87.8,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image1.png?download=true,"{""context"":""256K"",""effort"":""default (maximum reasoning setting)"",""harness"":""Mistral official maximum-reasoning comparison"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""pass@1; repeat count undisclosed"",""temperature"":""1.0"",""tools"":""none""}",False,official_model_card,verified,"Displayed exactly: '87.8'. Research observation: medium-image1.png:3:4:score:reported. Chart footnotes: *All model evaluations were run with maximum reasoning settings. | **Sonnet 4.6 encountered higher rates of reasoning truncation due to external API restrictions than other models, which affected its performance."
kimi-k2.5,tau3_airline,76.5,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image4.png?download=true,"{""context"":""256K"",""effort"":""default (maximum reasoning setting)"",""harness"":""Mistral \u03c4\u00b3 evaluation with GPT-5.2 low-effort user simulator"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""trials=4"",""temperature"":""1.0"",""tools"":""\u03c4\u00b3 domain tools and retrieval environment""}",False,official_model_card,verified,"Displayed exactly: '76.5'. Research observation: medium-image4.png:3:4:score:reported. Chart footnotes: *All model evaluations were run with maximum reasoning settings. | **Sonnet 4.6 encountered higher rates of reasoning truncation due to external API restrictions than other models, which affected its performance. | *** Self-reported | **** Self-reported, Mistral is using context management and a discard-all strategy at 100k tokens. | τ³ scores as reported by Sierra for Claude Sonnet 4.5 and Qwen3.5. Others with user simulator: gpt-5.2 with reasoning_effort: low. 4 trials. Banking domain evaluated with terminal- or embedding-based agentic search retrieval, only highest score is reported."
kimi-k2.5,tau3_banking,14.9,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image4.png?download=true,"{""context"":""256K"",""effort"":""default (maximum reasoning setting)"",""harness"":""Mistral \u03c4\u00b3 evaluation with GPT-5.2 low-effort user simulator"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""trials=4"",""temperature"":""1.0"",""tools"":""\u03c4\u00b3 Banking terminal or embedding retrieval; higher of the two reported""}",False,official_model_card,verified,"Displayed exactly: '14.9'. Research observation: medium-image4.png:5:4:score:reported. Chart footnotes: *All model evaluations were run with maximum reasoning settings. | **Sonnet 4.6 encountered higher rates of reasoning truncation due to external API restrictions than other models, which affected its performance. | *** Self-reported | **** Self-reported, Mistral is using context management and a discard-all strategy at 100k tokens. | τ³ scores as reported by Sierra for Claude Sonnet 4.5 and Qwen3.5. Others with user simulator: gpt-5.2 with reasoning_effort: low. 4 trials. Banking domain evaluated with terminal- or embedding-based agentic search retrieval, only highest score is reported."
kimi-k2.5,tau3_retail,72.8,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image4.png?download=true,"{""context"":""256K"",""effort"":""default (maximum reasoning setting)"",""harness"":""Mistral \u03c4\u00b3 evaluation with GPT-5.2 low-effort user simulator"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""trials=4"",""temperature"":""1.0"",""tools"":""\u03c4\u00b3 domain tools and retrieval environment""}",False,official_model_card,verified,"Displayed exactly: '72.8'. Research observation: medium-image4.png:4:4:score:reported. Chart footnotes: *All model evaluations were run with maximum reasoning settings. | **Sonnet 4.6 encountered higher rates of reasoning truncation due to external API restrictions than other models, which affected its performance. | *** Self-reported | **** Self-reported, Mistral is using context management and a discard-all strategy at 100k tokens. | τ³ scores as reported by Sierra for Claude Sonnet 4.5 and Qwen3.5. Others with user simulator: gpt-5.2 with reasoning_effort: low. 4 trials. Banking domain evaluated with terminal- or embedding-based agentic search retrieval, only highest score is reported."
kimi-k2.5,tau3_telecom,86.8,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image4.png?download=true,"{""context"":""256K"",""effort"":""default (maximum reasoning setting)"",""harness"":""Mistral \u03c4\u00b3 evaluation with GPT-5.2 low-effort user simulator"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default"",""sampling"":""trials=4"",""temperature"":""1.0"",""tools"":""\u03c4\u00b3 domain tools and retrieval environment""}",False,official_model_card,verified,"Displayed exactly: '86.8'. Research observation: medium-image4.png:2:4:score:reported. Chart footnotes: *All model evaluations were run with maximum reasoning settings. | **Sonnet 4.6 encountered higher rates of reasoning truncation due to external API restrictions than other models, which affected its performance. | *** Self-reported | **** Self-reported, Mistral is using context management and a discard-all strategy at 100k tokens. | τ³ scores as reported by Sierra for Claude Sonnet 4.5 and Qwen3.5. Others with user simulator: gpt-5.2 with reasoning_effort: low. 4 trials. Banking domain evaluated with terminal- or embedding-based agentic search retrieval, only highest score is reported."
magistral-medium-1.2,aa_lcr_mistral_custom_gpt_4_1_mini_middle_out,73.0,https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/resolve/a11f36bebf709121056b1dbcc943d1c6afbe494d/images/image2.png?download=true,"{""context"":""128k"",""effort"":""default"",""harness"":""Mistral custom AA-LCR with middle-out context handling"",""judge"":""gpt-4.1-mini-2025-04-14 equality judge"",""mode"":""thinking"",""prompt_style"":""official Magistral reasoning prompt"",""sampling"":""pass@1; repeat count undisclosed by Mistral"",""temperature"":""0.7; top_p=0.95 (family recommendation)"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly: '73'. Research observation: small-image2.png:1:2:score:reported.
magistral-medium-1.2,aime_2025,84.4,https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/resolve/a11f36bebf709121056b1dbcc943d1c6afbe494d/images/image2.png?download=true,"{""context"":""128k"",""effort"":""default"",""harness"":""Mistral official model-card evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official Magistral reasoning prompt"",""sampling"":""pass@1; repeat count undisclosed"",""temperature"":""0.7; top_p=0.95 (family recommendation)"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly: '84.4'. Research observation: small-image2.png:2:2:score:reported.
magistral-medium-1.2,browsecomp,10.0,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image2.png?download=true,"{""context"":""128k"",""effort"":""default"",""harness"":""Mistral official BrowseComp evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official Magistral reasoning prompt"",""sampling"":""source does not state trial count"",""temperature"":""0.7; top_p=0.95 (family recommendation)"",""tools"":""web search/browser environment""}",False,official_model_card,verified,Displayed exactly: '10.0'. Research observation: medium-image2.png:5:2:score:reported.
magistral-medium-1.2,collie,61.3,https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/resolve/a11f36bebf709121056b1dbcc943d1c6afbe494d/images/image2.png?download=true,"{""context"":""128k"",""effort"":""default"",""harness"":""Mistral official model-card evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official Magistral reasoning prompt"",""sampling"":""pass@1; repeat count undisclosed"",""temperature"":""0.7; top_p=0.95 (family recommendation)"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly: '61.3'. Research observation: small-image2.png:3:2:score:reported.
magistral-medium-1.2,tau3_airline,53.5,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image2.png?download=true,"{""context"":""128k"",""effort"":""default"",""harness"":""Mistral official \u03c4\u00b3 evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official Magistral reasoning prompt"",""sampling"":""source does not state trial count"",""temperature"":""0.7; top_p=0.95 (family recommendation)"",""tools"":""\u03c4\u00b3 domain tools and retrieval environment""}",False,official_model_card,verified,Displayed exactly: '53.5'. Research observation: medium-image2.png:2:2:score:reported.
magistral-medium-1.2,tau3_banking,7.7,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image2.png?download=true,"{""context"":""128k"",""effort"":""default"",""harness"":""Mistral official \u03c4\u00b3 evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official Magistral reasoning prompt"",""sampling"":""source does not state trial count"",""temperature"":""0.7; top_p=0.95 (family recommendation)"",""tools"":""\u03c4\u00b3 domain tools and retrieval environment""}",False,official_model_card,verified,Displayed exactly: '7.7'. Research observation: medium-image2.png:4:2:score:reported.
magistral-medium-1.2,tau3_retail,70.2,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image2.png?download=true,"{""context"":""128k"",""effort"":""default"",""harness"":""Mistral official \u03c4\u00b3 evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official Magistral reasoning prompt"",""sampling"":""source does not state trial count"",""temperature"":""0.7; top_p=0.95 (family recommendation)"",""tools"":""\u03c4\u00b3 domain tools and retrieval environment""}",False,official_model_card,verified,Displayed exactly: '70.2'. Research observation: medium-image2.png:3:2:score:reported.
magistral-medium-1.2,tau3_telecom,60.5,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image2.png?download=true,"{""context"":""128k"",""effort"":""default"",""harness"":""Mistral official \u03c4\u00b3 evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official Magistral reasoning prompt"",""sampling"":""source does not state trial count"",""temperature"":""0.7; top_p=0.95 (family recommendation)"",""tools"":""\u03c4\u00b3 domain tools and retrieval environment""}",False,official_model_card,verified,Displayed exactly: '60.5'. Research observation: medium-image2.png:1:2:score:reported.
magistral-small-1.2,aa_lcr_mistral_custom_gpt_4_1_mini_middle_out,27.0,https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/resolve/a11f36bebf709121056b1dbcc943d1c6afbe494d/images/image2.png?download=true,"{""context"":""128k"",""effort"":""default"",""harness"":""Mistral custom AA-LCR with middle-out context handling"",""judge"":""gpt-4.1-mini-2025-04-14 equality judge"",""mode"":""thinking"",""prompt_style"":""official Magistral reasoning system prompt"",""sampling"":""pass@1; repeat count undisclosed by Mistral"",""temperature"":""0.7; top_p=0.95; max_tokens=131072"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly: '27'. Research observation: small-image2.png:1:3:score:reported.
magistral-small-1.2,aime_2025,80.2,https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/resolve/a11f36bebf709121056b1dbcc943d1c6afbe494d/images/image2.png?download=true,"{""context"":""128k"",""effort"":""default"",""harness"":""Mistral official model-card evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official Magistral reasoning system prompt"",""sampling"":""pass@1; repeat count undisclosed"",""temperature"":""0.7; top_p=0.95; max_tokens=131072"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly: '80.2'. Research observation: small-image2.png:2:3:score:reported.
magistral-small-1.2,collie,60.3,https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/resolve/a11f36bebf709121056b1dbcc943d1c6afbe494d/images/image2.png?download=true,"{""context"":""128k"",""effort"":""default"",""harness"":""Mistral official model-card evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official Magistral reasoning system prompt"",""sampling"":""pass@1; repeat count undisclosed"",""temperature"":""0.7; top_p=0.95; max_tokens=131072"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly: '60.3'. Research observation: small-image2.png:3:3:score:reported.
mistral-large-3,ifbench,37.8,https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/resolve/a11f36bebf709121056b1dbcc943d1c6afbe494d/images/image3.png?download=true,"{""context"":""source does not state"",""effort"":""none"",""harness"":""Mistral official model-card evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official/default or source does not state"",""sampling"":""pass@1; repeat count undisclosed"",""temperature"":""source does not state"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly: '37.8'. Research observation: small-image3.png:3:4:score:reported.
mistral-large-3,mmmu_pro,54.0,https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/resolve/a11f36bebf709121056b1dbcc943d1c6afbe494d/images/image3.png?download=true,"{""context"":""source does not state"",""effort"":""none"",""harness"":""Mistral official model-card evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official/default or source does not state"",""sampling"":""pass@1; repeat count undisclosed"",""temperature"":""source does not state"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly: '54'. Research observation: small-image3.png:5:4:score:reported.
mistral-medium-3.1,arena_hard,67.3,https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/resolve/a11f36bebf709121056b1dbcc943d1c6afbe494d/images/image3.png?download=true,"{""context"":""128k"",""effort"":""none"",""harness"":""Mistral official model-card evaluation"",""judge"":""Arena-Hard automatic judge"",""mode"":""non-thinking"",""prompt_style"":""official/default"",""sampling"":""pass@1; repeat count undisclosed"",""temperature"":""undisclosed Mistral API default"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly: '67.3'. Research observation: small-image3.png:4:3:score:reported.
mistral-medium-3.1,browsecomp,7.8,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image2.png?download=true,"{""context"":""128k"",""effort"":""none"",""harness"":""Mistral official BrowseComp evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official/default"",""sampling"":""source does not state trial count"",""temperature"":""undisclosed Mistral API default"",""tools"":""web search/browser environment""}",False,official_model_card,verified,Displayed exactly: '7.8'. Research observation: medium-image2.png:5:4:score:reported.
mistral-medium-3.1,gpqa_diamond,65.7,https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/resolve/a11f36bebf709121056b1dbcc943d1c6afbe494d/images/image3.png?download=true,"{""context"":""128k"",""effort"":""none"",""harness"":""Mistral official model-card evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official/default"",""sampling"":""pass@1; repeat count undisclosed"",""temperature"":""undisclosed Mistral API default"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly: '65.7'. Research observation: small-image3.png:1:3:score:reported.
mistral-medium-3.1,ifbench,40.8,https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/resolve/a11f36bebf709121056b1dbcc943d1c6afbe494d/images/image3.png?download=true,"{""context"":""128k"",""effort"":""none"",""harness"":""Mistral official model-card evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official/default"",""sampling"":""pass@1; repeat count undisclosed"",""temperature"":""undisclosed Mistral API default"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly: '40.8'. Research observation: small-image3.png:3:3:score:reported.
mistral-medium-3.1,mmlu_pro,76.8,https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/resolve/a11f36bebf709121056b1dbcc943d1c6afbe494d/images/image3.png?download=true,"{""context"":""128k"",""effort"":""none"",""harness"":""Mistral official model-card evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official/default"",""sampling"":""pass@1; repeat count undisclosed"",""temperature"":""undisclosed Mistral API default"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly: '76.8'. Research observation: small-image3.png:2:3:score:reported.
mistral-medium-3.1,mmmu_pro,43.8,https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/resolve/a11f36bebf709121056b1dbcc943d1c6afbe494d/images/image3.png?download=true,"{""context"":""128k"",""effort"":""none"",""harness"":""Mistral official model-card evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official/default"",""sampling"":""pass@1; repeat count undisclosed"",""temperature"":""undisclosed Mistral API default"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly: '43.8'. Research observation: small-image3.png:5:3:score:reported.
mistral-medium-3.1,tau3_airline,41.5,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image2.png?download=true,"{""context"":""128k"",""effort"":""none"",""harness"":""Mistral official \u03c4\u00b3 evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official/default"",""sampling"":""source does not state trial count"",""temperature"":""undisclosed Mistral API default"",""tools"":""\u03c4\u00b3 domain tools and retrieval environment""}",False,official_model_card,verified,Displayed exactly: '41.5'. Research observation: medium-image2.png:2:4:score:reported.
mistral-medium-3.1,tau3_banking,5.7,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image2.png?download=true,"{""context"":""128k"",""effort"":""none"",""harness"":""Mistral official \u03c4\u00b3 evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official/default"",""sampling"":""source does not state trial count"",""temperature"":""undisclosed Mistral API default"",""tools"":""\u03c4\u00b3 domain tools and retrieval environment""}",False,official_model_card,verified,Displayed exactly: '5.7'. Research observation: medium-image2.png:4:4:score:reported.
mistral-medium-3.1,tau3_retail,64.3,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image2.png?download=true,"{""context"":""128k"",""effort"":""none"",""harness"":""Mistral official \u03c4\u00b3 evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official/default"",""sampling"":""source does not state trial count"",""temperature"":""undisclosed Mistral API default"",""tools"":""\u03c4\u00b3 domain tools and retrieval environment""}",False,official_model_card,verified,Displayed exactly: '64.3'. Research observation: medium-image2.png:3:4:score:reported.
mistral-medium-3.1,tau3_telecom,46.9,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image2.png?download=true,"{""context"":""128k"",""effort"":""none"",""harness"":""Mistral official \u03c4\u00b3 evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official/default"",""sampling"":""source does not state trial count"",""temperature"":""undisclosed Mistral API default"",""tools"":""\u03c4\u00b3 domain tools and retrieval environment""}",False,official_model_card,verified,Displayed exactly: '46.9'. Research observation: medium-image2.png:1:4:score:reported.
mistral-medium-3.5-128b,aime_2025,86.3,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image1.png?download=true,"{""context"":""256k"",""effort"":""high"",""harness"":""Mistral official maximum-reasoning comparison"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official/default"",""sampling"":""samples=16 (reported as avg@16)"",""temperature"":""0.7; top_p=0.95"",""tools"":""none""}",False,official_model_card,verified,"Displayed exactly: '86.3'. Research observation: medium-image1.png:1:1:score:reported. Chart footnotes: *All model evaluations were run with maximum reasoning settings. | **Sonnet 4.6 encountered higher rates of reasoning truncation due to external API restrictions than other models, which affected its performance."
mistral-medium-3.5-128b,beyond_aime,66.9,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image1.png?download=true,"{""context"":""256k"",""effort"":""high"",""harness"":""Mistral official maximum-reasoning comparison"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official/default"",""sampling"":""samples=16 (reported as avg@16)"",""temperature"":""0.7; top_p=0.95"",""tools"":""none""}",False,official_model_card,verified,"Displayed exactly: '66.9'. Research observation: medium-image1.png:4:1:score:reported. Chart footnotes: *All model evaluations were run with maximum reasoning settings. | **Sonnet 4.6 encountered higher rates of reasoning truncation due to external API restrictions than other models, which affected its performance."
mistral-medium-3.5-128b,browsecomp_cm,48.6,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image2.png?download=true,"{""context"":""256k"",""effort"":""high"",""harness"":""Mistral official BrowseComp evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official/default"",""sampling"":""source does not state trial count"",""temperature"":""0.7; top_p=0.95"",""tools"":""web search/browser with context management and discard-all at 100k tokens""}",False,official_model_card,verified,Displayed exactly: '48.6'. Research observation: medium-image2.png:5:1:score:reported. The companion chart footnote establishes that Medium 3.5 uses context management with discard-all at 100k tokens.
mistral-medium-3.5-128b,collie,95.8,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image1.png?download=true,"{""context"":""256k"",""effort"":""high"",""harness"":""Mistral official maximum-reasoning comparison"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official/default"",""sampling"":""pass@1; repeat count undisclosed"",""temperature"":""0.7; top_p=0.95"",""tools"":""none""}",False,official_model_card,verified,"Displayed exactly: '95.8'. Research observation: medium-image1.png:3:1:score:reported. Chart footnotes: *All model evaluations were run with maximum reasoning settings. | **Sonnet 4.6 encountered higher rates of reasoning truncation due to external API restrictions than other models, which affected its performance."
mistral-medium-3.5-128b,ifbench,69.0,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image1.png?download=true,"{""context"":""256k"",""effort"":""high"",""harness"":""Mistral official maximum-reasoning comparison"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official/default"",""sampling"":""pass@1; repeat count undisclosed"",""temperature"":""0.7; top_p=0.95"",""tools"":""none""}",False,official_model_card,verified,"Displayed exactly: '69.0'. Research observation: medium-image1.png:2:1:score:reported. Chart footnotes: *All model evaluations were run with maximum reasoning settings. | **Sonnet 4.6 encountered higher rates of reasoning truncation due to external API restrictions than other models, which affected its performance."
mistral-medium-3.5-128b,swe_bench_verified,77.6,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image3.png?download=true,"{""context"":""256k"",""effort"":""high"",""harness"":""Mistral self-reported SWE-bench Verified evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official/default"",""sampling"":""source does not state trial count"",""temperature"":""0.7; top_p=0.95"",""tools"":""agentic coding scaffold and SWE-bench verifier""}",False,official_model_card,verified,Displayed exactly: '77.6'. Research observation: medium-image3.png:1:1:score:reported.
mistral-medium-3.5-128b,tau3_airline,72.0,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image2.png?download=true,"{""context"":""256k"",""effort"":""high"",""harness"":""Mistral \u03c4\u00b3 evaluation with GPT-5.2 low-effort user simulator"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official/default"",""sampling"":""trials=4"",""temperature"":""0.7; top_p=0.95"",""tools"":""\u03c4\u00b3 domain tools and retrieval environment""}",False,official_model_card,verified,Displayed exactly: '72.0'. Research observation: medium-image2.png:2:1:score:reported.
mistral-medium-3.5-128b,tau3_banking,13.4,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image2.png?download=true,"{""context"":""256k"",""effort"":""high"",""harness"":""Mistral \u03c4\u00b3 evaluation with GPT-5.2 low-effort user simulator"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official/default"",""sampling"":""trials=4"",""temperature"":""0.7; top_p=0.95"",""tools"":""\u03c4\u00b3 Banking terminal or embedding retrieval; higher of the two reported""}",False,official_model_card,verified,Displayed exactly: '13.4'. Research observation: medium-image2.png:4:1:score:reported.
mistral-medium-3.5-128b,tau3_retail,76.1,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image2.png?download=true,"{""context"":""256k"",""effort"":""high"",""harness"":""Mistral \u03c4\u00b3 evaluation with GPT-5.2 low-effort user simulator"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official/default"",""sampling"":""trials=4"",""temperature"":""0.7; top_p=0.95"",""tools"":""\u03c4\u00b3 domain tools and retrieval environment""}",False,official_model_card,verified,Displayed exactly: '76.1'. Research observation: medium-image2.png:3:1:score:reported.
mistral-medium-3.5-128b,tau3_telecom,91.4,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image2.png?download=true,"{""context"":""256k"",""effort"":""high"",""harness"":""Mistral \u03c4\u00b3 evaluation with GPT-5.2 low-effort user simulator"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official/default"",""sampling"":""trials=4"",""temperature"":""0.7; top_p=0.95"",""tools"":""\u03c4\u00b3 domain tools and retrieval environment""}",False,official_model_card,verified,Displayed exactly: '91.4'. Research observation: medium-image2.png:1:1:score:reported.
mistral-small-3.2,arena_hard,43.1,https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/resolve/a11f36bebf709121056b1dbcc943d1c6afbe494d/images/image3.png?download=true,"{""context"":""128k"",""effort"":""none"",""harness"":""Mistral official model-card evaluation"",""judge"":""Arena-Hard automatic judge"",""mode"":""non-thinking"",""prompt_style"":""official SYSTEM_PROMPT.txt"",""sampling"":""pass@1; repeat count undisclosed"",""temperature"":""0.15 recommended"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly: '43.1'. Research observation: small-image3.png:4:2:score:reported.
mistral-small-3.2,gpqa_diamond,50.0,https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/resolve/a11f36bebf709121056b1dbcc943d1c6afbe494d/images/image3.png?download=true,"{""context"":""128k"",""effort"":""none"",""harness"":""Mistral official model-card evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official SYSTEM_PROMPT.txt"",""sampling"":""pass@1; repeat count undisclosed"",""temperature"":""0.15 recommended"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly: '50'. Research observation: small-image3.png:1:2:score:reported.
mistral-small-3.2,ifbench,34.0,https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/resolve/a11f36bebf709121056b1dbcc943d1c6afbe494d/images/image3.png?download=true,"{""context"":""128k"",""effort"":""none"",""harness"":""Mistral official model-card evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official SYSTEM_PROMPT.txt"",""sampling"":""pass@1; repeat count undisclosed"",""temperature"":""0.15 recommended"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly: '34'. Research observation: small-image3.png:3:2:score:reported.
mistral-small-3.2,mmlu_pro,69.0,https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/resolve/a11f36bebf709121056b1dbcc943d1c6afbe494d/images/image3.png?download=true,"{""context"":""128k"",""effort"":""none"",""harness"":""Mistral official model-card evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official SYSTEM_PROMPT.txt"",""sampling"":""pass@1; repeat count undisclosed"",""temperature"":""0.15 recommended"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly: '69'. Research observation: small-image3.png:2:2:score:reported.
mistral-small-3.2,mmmu_pro,49.1,https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/resolve/a11f36bebf709121056b1dbcc943d1c6afbe494d/images/image3.png?download=true,"{""context"":""128k"",""effort"":""none"",""harness"":""Mistral official model-card evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official SYSTEM_PROMPT.txt"",""sampling"":""pass@1; repeat count undisclosed"",""temperature"":""0.15 recommended"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly: '49.1'. Research observation: small-image3.png:5:2:score:reported.
mistral-small-4-119b-2603,aa_lcr_mistral_custom_gpt_4_1_mini_middle_out,71.2,https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/resolve/a11f36bebf709121056b1dbcc943d1c6afbe494d/images/image2.png?download=true,"{""context"":""256k"",""effort"":""high"",""harness"":""Mistral custom AA-LCR with middle-out context handling"",""judge"":""gpt-4.1-mini-2025-04-14 equality judge"",""mode"":""thinking"",""prompt_style"":""official/default"",""sampling"":""pass@1; repeat count undisclosed by Mistral"",""temperature"":""0.7"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly: '71.2'. Research observation: small-image2.png:1:1:score:high.
mistral-small-4-119b-2603,aime_2025,83.8,https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/resolve/a11f36bebf709121056b1dbcc943d1c6afbe494d/images/image2.png?download=true,"{""context"":""256k"",""effort"":""high"",""harness"":""Mistral official model-card evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official/default"",""sampling"":""pass@1; repeat count undisclosed"",""temperature"":""0.7"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly: '83.8'. Research observation: small-image2.png:2:1:score:high.
mistral-small-4-119b-2603,arena_hard,58.3,https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/resolve/a11f36bebf709121056b1dbcc943d1c6afbe494d/images/image3.png?download=true,"{""context"":""256k"",""effort"":""high"",""harness"":""Mistral official model-card evaluation"",""judge"":""Arena-Hard automatic judge"",""mode"":""thinking"",""prompt_style"":""official/default"",""sampling"":""pass@1; repeat count undisclosed"",""temperature"":""0.7"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly: '58.3'. Research observation: small-image3.png:4:1:score:reasoning.
mistral-small-4-119b-2603,browsecomp,21.3,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image2.png?download=true,"{""context"":""256k"",""effort"":""high"",""harness"":""Mistral official BrowseComp evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official/default"",""sampling"":""source does not state trial count"",""temperature"":""0.7"",""tools"":""web search/browser environment""}",False,official_model_card,verified,Displayed exactly: '21.3'. Research observation: medium-image2.png:5:3:score:reported.
mistral-small-4-119b-2603,collie,62.9,https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/resolve/a11f36bebf709121056b1dbcc943d1c6afbe494d/images/image2.png?download=true,"{""context"":""256k"",""effort"":""high"",""harness"":""Mistral official model-card evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official/default"",""sampling"":""pass@1; repeat count undisclosed"",""temperature"":""0.7"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly: '62.9'. Research observation: small-image2.png:3:1:score:high.
mistral-small-4-119b-2603,gpqa_diamond,71.2,https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/resolve/a11f36bebf709121056b1dbcc943d1c6afbe494d/images/image3.png?download=true,"{""context"":""256k"",""effort"":""high"",""harness"":""Mistral official model-card evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official/default"",""sampling"":""pass@1; repeat count undisclosed"",""temperature"":""0.7"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly: '71.2'. Research observation: small-image3.png:1:1:score:reasoning.
mistral-small-4-119b-2603,ifbench,48.0,https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/resolve/a11f36bebf709121056b1dbcc943d1c6afbe494d/images/image3.png?download=true,"{""context"":""256k"",""effort"":""high"",""harness"":""Mistral official model-card evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official/default"",""sampling"":""pass@1; repeat count undisclosed"",""temperature"":""0.7"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly: '48'. Research observation: small-image3.png:3:1:score:reasoning.
mistral-small-4-119b-2603,mmlu_pro,78.0,https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/resolve/a11f36bebf709121056b1dbcc943d1c6afbe494d/images/image3.png?download=true,"{""context"":""256k"",""effort"":""high"",""harness"":""Mistral official model-card evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official/default"",""sampling"":""pass@1; repeat count undisclosed"",""temperature"":""0.7"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly: '78'. Research observation: small-image3.png:2:1:score:reasoning.
mistral-small-4-119b-2603,mmmu_pro,60.0,https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/resolve/a11f36bebf709121056b1dbcc943d1c6afbe494d/images/image3.png?download=true,"{""context"":""256k"",""effort"":""high"",""harness"":""Mistral official model-card evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official/default"",""sampling"":""pass@1; repeat count undisclosed"",""temperature"":""0.7"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly: '60'. Research observation: small-image3.png:5:1:score:reasoning.
mistral-small-4-119b-2603,tau3_airline,38.5,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image2.png?download=true,"{""context"":""256k"",""effort"":""high"",""harness"":""Mistral official \u03c4\u00b3 evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official/default"",""sampling"":""source does not state trial count"",""temperature"":""0.7"",""tools"":""\u03c4\u00b3 domain tools and retrieval environment""}",False,official_model_card,verified,Displayed exactly: '38.5'. Research observation: medium-image2.png:2:3:score:reported.
mistral-small-4-119b-2603,tau3_banking,7.0,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image2.png?download=true,"{""context"":""256k"",""effort"":""high"",""harness"":""Mistral official \u03c4\u00b3 evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official/default"",""sampling"":""source does not state trial count"",""temperature"":""0.7"",""tools"":""\u03c4\u00b3 domain tools and retrieval environment""}",False,official_model_card,verified,Displayed exactly: '7.0'. Research observation: medium-image2.png:4:3:score:reported.
mistral-small-4-119b-2603,tau3_retail,67.8,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image2.png?download=true,"{""context"":""256k"",""effort"":""high"",""harness"":""Mistral official \u03c4\u00b3 evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official/default"",""sampling"":""source does not state trial count"",""temperature"":""0.7"",""tools"":""\u03c4\u00b3 domain tools and retrieval environment""}",False,official_model_card,verified,Displayed exactly: '67.8'. Research observation: medium-image2.png:3:3:score:reported.
mistral-small-4-119b-2603,tau3_telecom,47.1,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image2.png?download=true,"{""context"":""256k"",""effort"":""high"",""harness"":""Mistral official \u03c4\u00b3 evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official/default"",""sampling"":""source does not state trial count"",""temperature"":""0.7"",""tools"":""\u03c4\u00b3 domain tools and retrieval environment""}",False,official_model_card,verified,Displayed exactly: '47.1'. Research observation: medium-image2.png:1:3:score:reported.
qwen3-next-80b-a3b-instruct,aa_lcr_mistral_custom_gpt_4_1_mini_middle_out,62.0,https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/resolve/a11f36bebf709121056b1dbcc943d1c6afbe494d/images/aime.png?download=true,"{""context"":""262,144 native; approximately 1,010,000 with official extension"",""effort"":""none"",""harness"":""Mistral custom AA-LCR with middle-out context handling"",""judge"":""gpt-4.1-mini-2025-04-14 equality judge"",""mode"":""non-thinking"",""prompt_style"":""official; benchmark output-format prompts"",""sampling"":""pass@1; repeat count undisclosed by Mistral"",""temperature"":""0.7; top_p=0.8; top_k=20; min_p=0"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly: '62'. Research observation: small-aime.png:1:4:score:instruct. Chart footnotes: *Custom implementation of AA LCR using gpt-4.1-mini-2025-04-14 as a judge and a middle-out approach for models with shorter context lengths.
qwen3-next-80b-a3b-instruct,aime_2025,65.0,https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/resolve/a11f36bebf709121056b1dbcc943d1c6afbe494d/images/livecode.png?download=true,"{""context"":""262,144 native; approximately 1,010,000 with official extension"",""effort"":""none"",""harness"":""Mistral official model-card evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""official; benchmark output-format prompts"",""sampling"":""pass@1; repeat count undisclosed"",""temperature"":""0.7; top_p=0.8; top_k=20; min_p=0"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly: '65'. Research observation: small-livecode.png:1:4:score:instruct.
qwen3-next-80b-a3b-thinking,aa_lcr_mistral_custom_gpt_4_1_mini_middle_out,75.0,https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/resolve/a11f36bebf709121056b1dbcc943d1c6afbe494d/images/aime.png?download=true,"{""context"":""262,144 native; approximately 1,010,000 with official extension"",""effort"":""default"",""harness"":""Mistral custom AA-LCR with middle-out context handling"",""judge"":""gpt-4.1-mini-2025-04-14 equality judge"",""mode"":""thinking"",""prompt_style"":""official; benchmark output-format prompts"",""sampling"":""pass@1; repeat count undisclosed by Mistral"",""temperature"":""0.6; top_p=0.95; top_k=20; min_p=0"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly: '75'. Research observation: small-aime.png:1:4:score:reasoning. Chart footnotes: *Custom implementation of AA LCR using gpt-4.1-mini-2025-04-14 as a judge and a middle-out approach for models with shorter context lengths.
qwen3-next-80b-a3b-thinking,aime_2025,88.0,https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/resolve/a11f36bebf709121056b1dbcc943d1c6afbe494d/images/livecode.png?download=true,"{""context"":""262,144 native; approximately 1,010,000 with official extension"",""effort"":""default"",""harness"":""Mistral official model-card evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official; benchmark output-format prompts"",""sampling"":""pass@1; repeat count undisclosed"",""temperature"":""0.6; top_p=0.95; top_k=20; min_p=0"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly: '88'. Research observation: small-livecode.png:1:4:score:reasoning.
qwen3.5-122b-a10b,aa_lcr_mistral_custom_gpt_4_1_mini_middle_out,84.0,https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/resolve/a11f36bebf709121056b1dbcc943d1c6afbe494d/images/aime.png?download=true,"{""context"":""262,144 native"",""effort"":""none"",""harness"":""Mistral custom AA-LCR with middle-out context handling"",""judge"":""gpt-4.1-mini-2025-04-14 equality judge"",""mode"":""non-thinking"",""prompt_style"":""official"",""sampling"":""pass@1; repeat count undisclosed by Mistral"",""temperature"":""1.0; top_p=0.95; top_k=20; presence_penalty=1.5"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly: '84'. Research observation: small-aime.png:1:5:score:instruct. Chart footnotes: *Custom implementation of AA LCR using gpt-4.1-mini-2025-04-14 as a judge and a middle-out approach for models with shorter context lengths.
qwen3.5-122b-a10b,aime_2025,93.0,https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/resolve/a11f36bebf709121056b1dbcc943d1c6afbe494d/images/livecode.png?download=true,"{""context"":""262,144 native"",""effort"":""default"",""harness"":""Mistral official model-card evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""official"",""sampling"":""pass@1; repeat count undisclosed"",""temperature"":""1.0; top_p=0.95; top_k=20; presence_penalty=1.5"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly: '93'. Research observation: small-livecode.png:1:5:score:reasoning.
qwen3.5-397b,beyond_aime,72.3,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image1.png?download=true,"{""context"":""default (262K)"",""effort"":""default (maximum reasoning setting)"",""harness"":""Mistral official maximum-reasoning comparison"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default (thinking by default)"",""sampling"":""samples=16 (reported as avg@16)"",""temperature"":""default"",""tools"":""none""}",False,official_model_card,verified,"Displayed exactly: '72.3'. Research observation: medium-image1.png:4:5:score:reported. Chart footnotes: *All model evaluations were run with maximum reasoning settings. | **Sonnet 4.6 encountered higher rates of reasoning truncation due to external API restrictions than other models, which affected its performance."
qwen3.5-397b,collie,88.9,https://huggingface.co/mistralai/Mistral-Medium-3.5-128B/resolve/22b2b868a15677cfa6061277ed2f653d1349a9ab/images/image1.png?download=true,"{""context"":""default (262K)"",""effort"":""default (maximum reasoning setting)"",""harness"":""Mistral official maximum-reasoning comparison"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""default (thinking by default)"",""sampling"":""pass@1; repeat count undisclosed"",""temperature"":""default"",""tools"":""none""}",False,official_model_card,verified,"Displayed exactly: '88.9'. Research observation: medium-image1.png:3:5:score:reported. Chart footnotes: *All model evaluations were run with maximum reasoning settings. | **Sonnet 4.6 encountered higher rates of reasoning truncation due to external API restrictions than other models, which affected its performance."
gemma-3-1b-it,bfcl_v4,7.17,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 7.17. Official Liquid AI reported result.
gemma-3-1b-it,tau2_bench_avg,20.6,https://aypchzzf9pftwuto.public.blob.vercel-storage.com/lfm2_5_230m_benchmarks-EcJisF1zm7I7KiST75d7XsE2LwE0lD.png,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 20.6. Official Liquid AI reported result.
gemma-3-1b-it,tau2_bench_retail,6.43,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 6.43. Official Liquid AI reported result.
gemma-3-1b-it,tau2_bench_telecom,9.36,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 9.36. Official Liquid AI reported result.
gemma-4-26b-a4b,aa_omniscience_accuracy,14.37,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 14.37. Official Liquid AI reported result.
gemma-4-26b-a4b,aa_omniscience_index,-62.07,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as -62.07. Official Liquid AI reported result.
gemma-4-26b-a4b,aa_omniscience_non_hallucination,10.75,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 10.75. Official Liquid AI reported result.
gemma-4-26b-a4b,aime_2025,68.67,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 68.67. Official Liquid AI reported result.
gemma-4-26b-a4b,bfcl_v3,68.87,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 68.87. Official Liquid AI reported result.
gemma-4-26b-a4b,bfcl_v4,55.87,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 55.87. Official Liquid AI reported result.
gemma-4-26b-a4b,math_500,94.2,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 94.20. Official Liquid AI reported result.
gemma-4-26b-a4b,multi_if,82.06,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 82.06. Official Liquid AI reported result.
gemma-4-e2b,aa_omniscience_accuracy,7.0,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 7.00. Official Liquid AI reported result.
gemma-4-e2b,aa_omniscience_index,-72.0,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as -72. Official Liquid AI reported result.
gemma-4-e2b,aa_omniscience_non_hallucination,15.05,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 15.05. Official Liquid AI reported result.
gemma-4-e2b,aa_omniscience_public_accuracy,6.37,https://huggingface.co/LiquidAI/LFM2.5-2.6B/resolve/ab00687315bc1298e9d54e9c4b611dde9867ccc2/README.md,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly as 6.37. Official Liquid AI reported result.
gemma-4-e2b,aa_omniscience_public_index,-74.47,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as -74.47. Official Liquid AI reported result.
gemma-4-e2b,aa_omniscience_public_non_hallucination,13.67,https://huggingface.co/LiquidAI/LFM2.5-2.6B/resolve/ab00687315bc1298e9d54e9c4b611dde9867ccc2/README.md,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly as 13.67. Official Liquid AI reported result.
gemma-4-e2b,aime_2025,26.0,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 26. Official Liquid AI reported result.
gemma-4-e2b,bfcl_v3,56.44,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 56.44. Official Liquid AI reported result.
gemma-4-e2b,bfcl_v4,31.91,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 31.91. Official Liquid AI reported result.
gemma-4-e2b,browsecomp_plus_openclaw,8.31,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""BrowseComp-Plus fixed-corpus evaluation"",""judge"":""exact-answer benchmark evaluation"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""OpenClaw search/retrieval agent over fixed corpus""}",False,official_blog,verified,Displayed exactly as 8.31. Official Liquid AI reported result.
gemma-4-e2b,claw_eval_en_average,53.14,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Claw-Eval English tasks across all three public splits"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""Pass^3; N=3 independent trials"",""temperature"":""source does not state evaluation parameters"",""tools"":""OpenClaw agent environment""}",False,official_blog,verified,Displayed exactly as 53.14. Official Liquid AI reported result.
gemma-4-e2b,ifstruct_v1,64.85,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""IFStruct v1.0 validator without constrained decoding"",""judge"":""binary schema and structural validator"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 64.85. Official Liquid AI reported result.
gemma-4-e2b,math_500,64.0,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 64.00. Official Liquid AI reported result.
gemma-4-e2b,multi_if,69.7,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 69.70. Official Liquid AI reported result.
gemma-4-e2b,pinchbench,44.24,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""pass@1"",""temperature"":""source does not state evaluation parameters"",""tools"":""OpenClaw coding environment""}",False,official_blog,verified,Displayed exactly as 44.24. Official Liquid AI reported result.
gemma-4-e2b,tau3_banking,3.35,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 3.35. Official Liquid AI reported result.
gemma-4-e2b,tool_sandbox,52.4,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Apple ToolSandbox stateful conversational harness"",""judge"":""milestone and minefield similarity evaluation"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 52.40. Official Liquid AI reported result.
gemma-4-e4b,aa_omniscience_accuracy,8.1,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 8.10. Official Liquid AI reported result.
gemma-4-e4b,aa_omniscience_index,-50.67,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as -50.67. Official Liquid AI reported result.
gemma-4-e4b,aa_omniscience_non_hallucination,36.06,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 36.06. Official Liquid AI reported result.
gemma-4-e4b,aa_omniscience_public_accuracy,8.33,https://huggingface.co/LiquidAI/LFM2.5-2.6B/resolve/ab00687315bc1298e9d54e9c4b611dde9867ccc2/README.md,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly as 8.33. Official Liquid AI reported result.
gemma-4-e4b,aa_omniscience_public_index,-49.03,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as -49.03. Official Liquid AI reported result.
gemma-4-e4b,aa_omniscience_public_non_hallucination,37.42,https://huggingface.co/LiquidAI/LFM2.5-2.6B/resolve/ab00687315bc1298e9d54e9c4b611dde9867ccc2/README.md,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly as 37.42. Official Liquid AI reported result.
gemma-4-e4b,aime_2025,34.33,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 34.33. Official Liquid AI reported result.
gemma-4-e4b,bfcl_v3,57.31,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 57.31. Official Liquid AI reported result.
gemma-4-e4b,bfcl_v4,33.92,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 33.92. Official Liquid AI reported result.
gemma-4-e4b,browsecomp_plus_openclaw,15.9,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""BrowseComp-Plus fixed-corpus evaluation"",""judge"":""exact-answer benchmark evaluation"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""OpenClaw search/retrieval agent over fixed corpus""}",False,official_blog,verified,Displayed exactly as 15.90. Official Liquid AI reported result.
gemma-4-e4b,claw_eval_en_average,58.02,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Claw-Eval English tasks across all three public splits"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""Pass^3; N=3 independent trials"",""temperature"":""source does not state evaluation parameters"",""tools"":""OpenClaw agent environment""}",False,official_blog,verified,Displayed exactly as 58.02. Official Liquid AI reported result.
gemma-4-e4b,ifstruct_v1,76.65,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""IFStruct v1.0 validator without constrained decoding"",""judge"":""binary schema and structural validator"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 76.65. Official Liquid AI reported result.
gemma-4-e4b,math_500,65.0,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 65.00. Official Liquid AI reported result.
gemma-4-e4b,multi_if,77.58,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 77.58. Official Liquid AI reported result.
gemma-4-e4b,pinchbench,55.09,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""pass@1"",""temperature"":""source does not state evaluation parameters"",""tools"":""OpenClaw coding environment""}",False,official_blog,verified,Displayed exactly as 55.09. Official Liquid AI reported result.
gemma-4-e4b,tau3_banking,4.12,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 4.12. Official Liquid AI reported result.
gemma-4-e4b,tool_sandbox,65.0,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Apple ToolSandbox stateful conversational harness"",""judge"":""milestone and minefield similarity evaluation"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 65.00. Official Liquid AI reported result.
gpt-oss-20b,aa_omniscience_accuracy,14.57,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 14.57. Official Liquid AI reported result.
gpt-oss-20b,aa_omniscience_index,-49.17,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as -49.17. Official Liquid AI reported result.
gpt-oss-20b,aa_omniscience_non_hallucination,24.5,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 24.50. Official Liquid AI reported result.
gpt-oss-20b,aime_2026,68.67,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 68.67. Official Liquid AI reported result.
gpt-oss-20b,bfcl_v3,62.52,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 62.52. Official Liquid AI reported result.
gpt-oss-20b,bfcl_v4,49.88,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 49.88. Official Liquid AI reported result.
gpt-oss-20b,ifbench,58.65,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 58.65. Official Liquid AI reported result.
gpt-oss-20b,multi_if,76.64,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 76.64. Official Liquid AI reported result.
gpt-oss-20b,tau2_bench_retail,53.51,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 53.51. Official Liquid AI reported result.
gpt-oss-20b,tau2_bench_telecom,57.24,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 57.24. Official Liquid AI reported result.
granite-4.0-350m,bfcl_v3,39.58,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 39.58. Official Liquid AI reported result.
granite-4.0-350m,bfcl_v4,13.73,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 13.73. Official Liquid AI reported result.
granite-4.0-350m,gpqa_diamond,25.91,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 25.91. Official Liquid AI reported result.
granite-4.0-350m,ifbench,15.98,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 15.98. Official Liquid AI reported result.
granite-4.0-350m,ifeval,53.48,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 53.48. Official Liquid AI reported result.
granite-4.0-350m,mmlu_pro,12.84,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 12.84. Official Liquid AI reported result.
granite-4.0-350m,multi_if,24.21,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 24.21. Official Liquid AI reported result.
granite-4.0-350m,tau2_bench_retail,6.14,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 6.14. Official Liquid AI reported result.
granite-4.0-350m,tau2_bench_telecom,2.92,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 2.92. Official Liquid AI reported result.
granite-4.0-h-350m,bfcl_v3,43.07,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 43.07. Official Liquid AI reported result.
granite-4.0-h-350m,bfcl_v4,13.28,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 13.28. Official Liquid AI reported result.
granite-4.0-h-350m,gpqa_diamond,22.32,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 22.32. Official Liquid AI reported result.
granite-4.0-h-350m,ifbench,17.22,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 17.22. Official Liquid AI reported result.
granite-4.0-h-350m,ifeval,61.27,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 61.27. Official Liquid AI reported result.
granite-4.0-h-350m,mmlu_pro,13.14,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 13.14. Official Liquid AI reported result.
granite-4.0-h-350m,multi_if,28.7,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 28.70. Official Liquid AI reported result.
granite-4.0-h-350m,tau2_bench_avg,18.85,https://aypchzzf9pftwuto.public.blob.vercel-storage.com/lfm2_5_230m_benchmarks-EcJisF1zm7I7KiST75d7XsE2LwE0lD.png,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 18.85. Official Liquid AI reported result.
granite-4.0-h-350m,tau2_bench_retail,6.14,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 6.14. Official Liquid AI reported result.
granite-4.0-h-350m,tau2_bench_telecom,13.74,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 13.74. Official Liquid AI reported result.
granite-4.0-h-tiny,aa_omniscience_accuracy,9.37,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 9.37. Official Liquid AI reported result.
granite-4.0-h-tiny,aa_omniscience_index,-75.5,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as -75.50. Official Liquid AI reported result.
granite-4.0-h-tiny,aa_omniscience_non_hallucination,6.38,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 6.38. Official Liquid AI reported result.
granite-4.0-h-tiny,aime_2025,4.93,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 4.93. Official Liquid AI reported result.
granite-4.0-h-tiny,aime_2026,3.33,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 3.33. Official Liquid AI reported result.
granite-4.0-h-tiny,bfcl_v3,56.89,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 56.89. Official Liquid AI reported result.
granite-4.0-h-tiny,bfcl_v4,28.52,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 28.52. Official Liquid AI reported result.
granite-4.0-h-tiny,ifbench,21.28,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 21.28. Official Liquid AI reported result.
granite-4.0-h-tiny,ifeval,82.23,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 82.23. Official Liquid AI reported result.
granite-4.0-h-tiny,math_500,59.2,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 59.20. Official Liquid AI reported result.
granite-4.0-h-tiny,multi_if,59.0,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 59.00. Official Liquid AI reported result.
granite-4.0-h-tiny,tau2_bench_retail,18.42,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 18.42. Official Liquid AI reported result.
granite-4.0-h-tiny,tau2_bench_telecom,16.67,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 16.67. Official Liquid AI reported result.
lfm2-350m,bfcl_v3,22.95,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 22.95. Official Liquid AI reported result.
lfm2-350m,bfcl_v4,12.29,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 12.29. Official Liquid AI reported result.
lfm2-350m,gpqa_diamond,27.58,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 27.58. Official Liquid AI reported result.
lfm2-350m,ifbench,18.2,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 18.20. Official Liquid AI reported result.
lfm2-350m,ifeval,64.96,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 64.96. Official Liquid AI reported result.
lfm2-350m,mmlu_pro,19.29,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 19.29. Official Liquid AI reported result.
lfm2-350m,multi_if,32.92,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 32.92. Official Liquid AI reported result.
lfm2-350m,tau2_bench_avg,20.13,https://aypchzzf9pftwuto.public.blob.vercel-storage.com/lfm2_5_230m_benchmarks-EcJisF1zm7I7KiST75d7XsE2LwE0lD.png,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 20.13. Official Liquid AI reported result.
lfm2-350m,tau2_bench_retail,5.56,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 5.56. Official Liquid AI reported result.
lfm2-350m,tau2_bench_telecom,10.82,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 10.82. Official Liquid AI reported result.
lfm2-8b-a1b,aa_omniscience_accuracy,7.33,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""non-thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 7.33. Official Liquid AI reported result.
lfm2-8b-a1b,aa_omniscience_index,-78.42,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""non-thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as -78.42. Official Liquid AI reported result.
lfm2-8b-a1b,aa_omniscience_non_hallucination,7.46,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""non-thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 7.46. Official Liquid AI reported result.
lfm2-8b-a1b,aime_2025,20.0,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 20.00. Official Liquid AI reported result.
lfm2-8b-a1b,bfcl_v3,45.07,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 45.07. Official Liquid AI reported result.
lfm2-8b-a1b,bfcl_v4,25.52,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 25.52. Official Liquid AI reported result.
lfm2-8b-a1b,ifbench,26.0,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 26.00. Official Liquid AI reported result.
lfm2-8b-a1b,ifeval,79.44,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 79.44. Official Liquid AI reported result.
lfm2-8b-a1b,math_500,74.8,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 74.80. Official Liquid AI reported result.
lfm2-8b-a1b,multi_if,58.54,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 58.54. Official Liquid AI reported result.
lfm2-8b-a1b,tau2_bench_retail,7.02,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 7.02. Official Liquid AI reported result.
lfm2-8b-a1b,tau2_bench_telecom,13.6,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 13.60. Official Liquid AI reported result.
lfm2.5-2.6b,aa_omniscience_public_accuracy,8.13,https://huggingface.co/LiquidAI/LFM2.5-2.6B/resolve/ab00687315bc1298e9d54e9c4b611dde9867ccc2/README.md,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly as 8.13. Official Liquid AI reported result.
lfm2.5-2.6b,aa_omniscience_public_index,-29.5,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as -29.50. Official Liquid AI reported result.
lfm2.5-2.6b,aa_omniscience_public_non_hallucination,59.04,https://huggingface.co/LiquidAI/LFM2.5-2.6B/resolve/ab00687315bc1298e9d54e9c4b611dde9867ccc2/README.md,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly as 59.04. Official Liquid AI reported result.
lfm2.5-2.6b,aime_2025,51.87,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 51.87. Official Liquid AI reported result.
lfm2.5-2.6b,bfcl_v4,56.88,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 56.88. Official Liquid AI reported result.
lfm2.5-2.6b,browsecomp_plus_openclaw,26.89,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""BrowseComp-Plus fixed-corpus evaluation"",""judge"":""exact-answer benchmark evaluation"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""OpenClaw search/retrieval agent over fixed corpus""}",False,official_blog,verified,Displayed exactly as 26.89. Official Liquid AI reported result.
lfm2.5-2.6b,claw_eval_en_average,62.85,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Claw-Eval English tasks across all three public splits"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""Pass^3; N=3 independent trials"",""temperature"":""source does not state evaluation parameters"",""tools"":""OpenClaw agent environment""}",False,official_blog,verified,Displayed exactly as 62.85. Official Liquid AI reported result.
lfm2.5-2.6b,ifbench,59.17,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 59.17. Official Liquid AI reported result.
lfm2.5-2.6b,ifstruct_v1,85.49,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""IFStruct v1.0 validator without constrained decoding"",""judge"":""binary schema and structural validator"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 85.49. Official Liquid AI reported result.
lfm2.5-2.6b,livecodebench_v6,59.41,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 59.41. Official Liquid AI reported result.
lfm2.5-2.6b,multi_if,80.07,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 80.07. Official Liquid AI reported result.
lfm2.5-2.6b,pinchbench,68.22,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""pass@1"",""temperature"":""source does not state evaluation parameters"",""tools"":""OpenClaw coding environment""}",False,official_blog,verified,Displayed exactly as 68.22. Official Liquid AI reported result.
lfm2.5-2.6b,tau3_banking,5.67,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 5.67. Official Liquid AI reported result.
lfm2.5-2.6b,tool_sandbox,77.83,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Apple ToolSandbox stateful conversational harness"",""judge"":""milestone and minefield similarity evaluation"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 77.83. Official Liquid AI reported result.
lfm2.5-230m,bfcl_v3,43.26,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 43.26. Official Liquid AI reported result.
lfm2.5-230m,bfcl_v4,21.03,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 21.03. Official Liquid AI reported result.
lfm2.5-230m,gpqa_diamond,25.41,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 25.41. Official Liquid AI reported result.
lfm2.5-230m,ifbench,38.4,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 38.40. Official Liquid AI reported result.
lfm2.5-230m,ifeval,71.71,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 71.71. Official Liquid AI reported result.
lfm2.5-230m,mmlu_pro,20.25,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 20.25. Official Liquid AI reported result.
lfm2.5-230m,multi_if,37.7,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 37.70. Official Liquid AI reported result.
lfm2.5-230m,tau2_bench_avg,16.98,https://aypchzzf9pftwuto.public.blob.vercel-storage.com/lfm2_5_230m_benchmarks-EcJisF1zm7I7KiST75d7XsE2LwE0lD.png,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 16.98. Official Liquid AI reported result.
lfm2.5-230m,tau2_bench_retail,13.68,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 13.68. Official Liquid AI reported result.
lfm2.5-230m,tau2_bench_telecom,5.26,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 5.26. Official Liquid AI reported result.
lfm2.5-350m,bfcl_v3,44.11,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 44.11. Official Liquid AI reported result.
lfm2.5-350m,bfcl_v4,21.86,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 21.86. Official Liquid AI reported result.
lfm2.5-350m,gpqa_diamond,30.64,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 30.64. Official Liquid AI reported result.
lfm2.5-350m,ifbench,40.69,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 40.69. Official Liquid AI reported result.
lfm2.5-350m,ifeval,76.96,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 76.96. Official Liquid AI reported result.
lfm2.5-350m,mmlu_pro,20.01,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 20.01. Official Liquid AI reported result.
lfm2.5-350m,multi_if,44.92,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 44.92. Official Liquid AI reported result.
lfm2.5-350m,tau2_bench_avg,23.09,https://aypchzzf9pftwuto.public.blob.vercel-storage.com/lfm2_5_230m_benchmarks-EcJisF1zm7I7KiST75d7XsE2LwE0lD.png,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 23.09. Official Liquid AI reported result.
lfm2.5-350m,tau2_bench_retail,17.84,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 17.84. Official Liquid AI reported result.
lfm2.5-350m,tau2_bench_telecom,18.86,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 18.86. Official Liquid AI reported result.
lfm2.5-8b-a1b,aa_omniscience_accuracy,8.67,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 8.67. Official Liquid AI reported result.
lfm2.5-8b-a1b,aa_omniscience_index,-24.7,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as -24.70. Official Liquid AI reported result.
lfm2.5-8b-a1b,aa_omniscience_non_hallucination,63.47,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 63.47. Official Liquid AI reported result.
lfm2.5-8b-a1b,aime_2025,42.53,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 42.53. Official Liquid AI reported result.
lfm2.5-8b-a1b,aime_2026,50.0,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 50.00. Official Liquid AI reported result.
lfm2.5-8b-a1b,bfcl_v3,64.79,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 64.79. Official Liquid AI reported result.
lfm2.5-8b-a1b,bfcl_v4,49.73,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 49.73. Official Liquid AI reported result.
lfm2.5-8b-a1b,ifbench,56.47,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 56.47. Official Liquid AI reported result.
lfm2.5-8b-a1b,ifeval,91.84,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 91.84. Official Liquid AI reported result.
lfm2.5-8b-a1b,math_500,88.76,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 88.76. Official Liquid AI reported result.
lfm2.5-8b-a1b,multi_if,79.93,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 79.93. Official Liquid AI reported result.
lfm2.5-8b-a1b,tau2_bench_retail,39.82,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 39.82. Official Liquid AI reported result.
lfm2.5-8b-a1b,tau2_bench_telecom,88.07,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 88.07. Official Liquid AI reported result.
qwen3-30b-a3b-thinking-2507,aa_omniscience_accuracy,18.8,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 18.80. Official Liquid AI reported result.
qwen3-30b-a3b-thinking-2507,aa_omniscience_index,-51.31,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as -51.31. Official Liquid AI reported result.
qwen3-30b-a3b-thinking-2507,aa_omniscience_non_hallucination,13.87,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 13.87. Official Liquid AI reported result.
qwen3-30b-a3b-thinking-2507,aime_2025,71.67,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 71.67. Official Liquid AI reported result.
qwen3-30b-a3b-thinking-2507,aime_2026,66.67,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 66.67. Official Liquid AI reported result.
qwen3-30b-a3b-thinking-2507,bfcl_v3,73.39,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 73.39. Official Liquid AI reported result.
qwen3-30b-a3b-thinking-2507,bfcl_v4,50.53,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 50.53. Official Liquid AI reported result.
qwen3-30b-a3b-thinking-2507,ifbench,51.11,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 51.11. Official Liquid AI reported result.
qwen3-30b-a3b-thinking-2507,ifeval,90.82,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 90.82. Official Liquid AI reported result.
qwen3-30b-a3b-thinking-2507,math_500,86.48,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 86.48. Official Liquid AI reported result.
qwen3-30b-a3b-thinking-2507,multi_if,79.04,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 79.04. Official Liquid AI reported result.
qwen3-30b-a3b-thinking-2507,tau2_bench_retail,56.14,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 56.14. Official Liquid AI reported result.
qwen3-30b-a3b-thinking-2507,tau2_bench_telecom,21.93,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 21.93. Official Liquid AI reported result.
qwen3.5-0.8b-instruct,bfcl_v3,35.08,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 35.08. Official Liquid AI reported result.
qwen3.5-0.8b-instruct,bfcl_v4,18.7,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 18.70. Official Liquid AI reported result.
qwen3.5-0.8b-instruct,gpqa_diamond,27.41,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 27.41. Official Liquid AI reported result.
qwen3.5-0.8b-instruct,ifbench,22.87,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 22.87. Official Liquid AI reported result.
qwen3.5-0.8b-instruct,ifeval,59.94,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 59.94. Official Liquid AI reported result.
qwen3.5-0.8b-instruct,mmlu_pro,37.42,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 37.42. Official Liquid AI reported result.
qwen3.5-0.8b-instruct,multi_if,41.68,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 41.68. Official Liquid AI reported result.
qwen3.5-0.8b-instruct,tau2_bench_avg,20.9,https://aypchzzf9pftwuto.public.blob.vercel-storage.com/lfm2_5_230m_benchmarks-EcJisF1zm7I7KiST75d7XsE2LwE0lD.png,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 20.9. Official Liquid AI reported result.
qwen3.5-0.8b-instruct,tau2_bench_retail,6.14,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 6.14. Official Liquid AI reported result.
qwen3.5-0.8b-instruct,tau2_bench_telecom,12.57,https://www.liquid.ai/blog/lfm2-5-230m,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""non-thinking"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 12.57. Official Liquid AI reported result.
qwen3.5-4b,aa_omniscience_accuracy,17.2,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 17.20. Official Liquid AI reported result.
qwen3.5-4b,aa_omniscience_index,-51.53,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as -51.53. Official Liquid AI reported result.
qwen3.5-4b,aa_omniscience_non_hallucination,16.99,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 16.99. Official Liquid AI reported result.
qwen3.5-4b,aa_omniscience_public_accuracy,17.63,https://huggingface.co/LiquidAI/LFM2.5-2.6B/resolve/ab00687315bc1298e9d54e9c4b611dde9867ccc2/README.md,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly as 17.63. Official Liquid AI reported result.
qwen3.5-4b,aa_omniscience_public_index,-54.3,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as -54.30. Official Liquid AI reported result.
qwen3.5-4b,aa_omniscience_public_non_hallucination,12.66,https://huggingface.co/LiquidAI/LFM2.5-2.6B/resolve/ab00687315bc1298e9d54e9c4b611dde9867ccc2/README.md,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly as 12.66. Official Liquid AI reported result.
qwen3.5-4b,aime_2025,54.28,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 54.28. Official Liquid AI reported result.
qwen3.5-4b,aime_2026,58.33,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 58.33. Official Liquid AI reported result.
qwen3.5-4b,bfcl_v3,71.06,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 71.06. Official Liquid AI reported result.
qwen3.5-4b,bfcl_v4,54.01,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 54.01. Official Liquid AI reported result.
qwen3.5-4b,browsecomp_plus_openclaw,24.46,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""BrowseComp-Plus fixed-corpus evaluation"",""judge"":""exact-answer benchmark evaluation"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""OpenClaw search/retrieval agent over fixed corpus""}",False,official_blog,verified,Displayed exactly as 24.46. Official Liquid AI reported result.
qwen3.5-4b,claw_eval_en_average,62.28,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Claw-Eval English tasks across all three public splits"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""Pass^3; N=3 independent trials"",""temperature"":""source does not state evaluation parameters"",""tools"":""OpenClaw agent environment""}",False,official_blog,verified,Displayed exactly as 62.28. Official Liquid AI reported result.
qwen3.5-4b,ifbench,50.38,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 50.38. Official Liquid AI reported result.
qwen3.5-4b,ifeval,87.8,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 87.80. Official Liquid AI reported result.
qwen3.5-4b,ifstruct_v1,36.25,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""IFStruct v1.0 validator without constrained decoding"",""judge"":""binary schema and structural validator"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 36.25. Official Liquid AI reported result.
qwen3.5-4b,livecodebench_v6,60.85,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 60.85. Official Liquid AI reported result.
qwen3.5-4b,math_500,80.76,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 80.76. Official Liquid AI reported result.
qwen3.5-4b,multi_if,67.43,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 67.43. Official Liquid AI reported result.
qwen3.5-4b,pinchbench,71.26,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""pass@1"",""temperature"":""source does not state evaluation parameters"",""tools"":""OpenClaw coding environment""}",False,official_blog,verified,Displayed exactly as 71.26. Official Liquid AI reported result.
qwen3.5-4b,tau2_bench_retail,71.93,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 71.93. Official Liquid AI reported result.
qwen3.5-4b,tau2_bench_telecom,87.72,https://www.liquid.ai/blog/lfm2-5-8b-a1b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 87.72. Official Liquid AI reported result.
qwen3.5-4b,tau3_banking,5.45,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 5.45. Official Liquid AI reported result.
qwen3.5-4b,tool_sandbox,75.55,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Apple ToolSandbox stateful conversational harness"",""judge"":""milestone and minefield similarity evaluation"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 75.55. Official Liquid AI reported result.
qwen3.5-9b,aa_omniscience_public_accuracy,21.3,https://huggingface.co/LiquidAI/LFM2.5-2.6B/resolve/ab00687315bc1298e9d54e9c4b611dde9867ccc2/README.md,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly as 21.30. Official Liquid AI reported result.
qwen3.5-9b,aa_omniscience_public_index,-50.43,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as -50.43. Official Liquid AI reported result.
qwen3.5-9b,aa_omniscience_public_non_hallucination,8.84,https://huggingface.co/LiquidAI/LFM2.5-2.6B/resolve/ab00687315bc1298e9d54e9c4b611dde9867ccc2/README.md,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Artificial Analysis Omniscience evaluation reported by Liquid"",""judge"":""official answer and abstention classification"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly as 8.84. Official Liquid AI reported result.
qwen3.5-9b,aime_2025,56.07,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 56.07. Official Liquid AI reported result.
qwen3.5-9b,bfcl_v4,60.13,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 60.13. Official Liquid AI reported result.
qwen3.5-9b,browsecomp_plus_openclaw,27.23,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""BrowseComp-Plus fixed-corpus evaluation"",""judge"":""exact-answer benchmark evaluation"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""OpenClaw search/retrieval agent over fixed corpus""}",False,official_blog,verified,Displayed exactly as 27.23. Official Liquid AI reported result.
qwen3.5-9b,claw_eval_en_average,66.53,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Claw-Eval English tasks across all three public splits"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""Pass^3; N=3 independent trials"",""temperature"":""source does not state evaluation parameters"",""tools"":""OpenClaw agent environment""}",False,official_blog,verified,Displayed exactly as 66.53. Official Liquid AI reported result.
qwen3.5-9b,ifbench,56.47,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 56.47. Official Liquid AI reported result.
qwen3.5-9b,ifstruct_v1,78.5,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""IFStruct v1.0 validator without constrained decoding"",""judge"":""binary schema and structural validator"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 78.50. Official Liquid AI reported result.
qwen3.5-9b,livecodebench_v6,69.86,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 69.86. Official Liquid AI reported result.
qwen3.5-9b,multi_if,62.55,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""none""}",False,official_blog,verified,Displayed exactly as 62.55. Official Liquid AI reported result.
qwen3.5-9b,pinchbench,71.45,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""pass@1"",""temperature"":""source does not state evaluation parameters"",""tools"":""OpenClaw coding environment""}",False,official_blog,verified,Displayed exactly as 71.45. Official Liquid AI reported result.
qwen3.5-9b,tau3_banking,5.15,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Liquid AI official comparative evaluation"",""judge"":""benchmark-specified"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 5.15. Official Liquid AI reported result.
qwen3.5-9b,tool_sandbox,76.44,https://www.liquid.ai/blog/lfm2-5-2-6b,"{""context"":""source does not state evaluation context"",""effort"":""source does not state"",""harness"":""Apple ToolSandbox stateful conversational harness"",""judge"":""milestone and minefield similarity evaluation"",""mode"":""source/model default"",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state evaluation parameters"",""tools"":""benchmark-provided tools""}",False,official_blog,verified,Displayed exactly as 76.44. Official Liquid AI reported result.
claude-haiku-4.5,advancedif_average,56.9,https://microsoft.ai/pdf/MAI-Code-1-Flash-Model-Card.PDF,"{""context"":""256K"",""effort"":""adaptive solution length control"",""harness"":""Microsoft MAI-Code-1-Flash comparative evaluation"",""judge"":""benchmark-specific"",""mode"":""adaptive reasoning"",""notes"":""Official instruction-following or agentic-tool-use score; benchmark=advancedif_average."",""prompt_style"":""source does not state"",""sampling"":""single/multi-turn average where applicable"",""temperature"":""source does not state"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly as 56.9. Official Microsoft-reported result.
claude-haiku-4.5,aime_2026,83.3,https://microsoft.ai/pdf/MAI-Code-1-Flash-Model-Card.PDF,"{""context"":""256K model context; per-task usage reported separately"",""effort"":""adaptive solution length control"",""harness"":""Microsoft MAI-Code-1-Flash comparative evaluation"",""judge"":""benchmark-specific"",""mode"":""adaptive reasoning"",""notes"":""Accuracy column from the official model-card/release table; benchmark=aime_2026."",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly as 83.3. Official Microsoft-reported result.
claude-haiku-4.5,amo_bench,16.0,https://microsoft.ai/pdf/MAI-Code-1-Flash-Model-Card.PDF,"{""context"":""256K model context; per-task usage reported separately"",""effort"":""adaptive solution length control"",""harness"":""Microsoft MAI-Code-1-Flash comparative evaluation"",""judge"":""benchmark-specific"",""mode"":""adaptive reasoning"",""notes"":""Accuracy column from the official model-card/release table; benchmark=amo_bench."",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly as 16.0. Official Microsoft-reported result.
claude-haiku-4.5,artifactsbench,36.6,https://microsoft.ai/pdf/MAI-Code-1-Flash-Model-Card.PDF,"{""context"":""256K model context; per-task usage reported separately"",""effort"":""adaptive solution length control"",""harness"":""Microsoft MAI-Code-1-Flash comparative evaluation"",""judge"":""benchmark-specific"",""mode"":""adaptive reasoning"",""notes"":""Accuracy column from the official model-card/release table; benchmark=artifactsbench."",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly as 36.6. Official Microsoft-reported result.
claude-haiku-4.5,frontiermath,2.8,https://microsoft.ai/pdf/MAI-Code-1-Flash-Model-Card.PDF,"{""context"":""256K model context; per-task usage reported separately"",""effort"":""adaptive solution length control"",""harness"":""Microsoft MAI-Code-1-Flash comparative evaluation"",""judge"":""benchmark-specific"",""mode"":""adaptive reasoning"",""notes"":""Accuracy column from the official model-card/release table; benchmark=frontiermath."",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly as 2.8. Official Microsoft-reported result.
claude-haiku-4.5,ifbench_single_multiturn_average,46.1,https://microsoft.ai/pdf/MAI-Code-1-Flash-Model-Card.PDF,"{""context"":""256K"",""effort"":""adaptive solution length control"",""harness"":""Microsoft MAI-Code-1-Flash comparative evaluation"",""judge"":""benchmark-specific"",""mode"":""adaptive reasoning"",""notes"":""Official instruction-following or agentic-tool-use score; benchmark=ifbench_single_multiturn_average."",""prompt_style"":""source does not state"",""sampling"":""single/multi-turn average where applicable"",""temperature"":""source does not state"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly as 46.1. Official Microsoft-reported result.
claude-haiku-4.5,swe_bench_multilingual,62.7,https://microsoft.ai/pdf/MAI-Code-1-Flash-Model-Card.PDF,"{""context"":""256K"",""effort"":""adaptive solution length control"",""harness"":""GitHub Copilot VS Code production harness"",""judge"":""benchmark-specified"",""mode"":""adaptive reasoning"",""notes"":""Coding-focused 137B-total / 5B-active model. Core coding scores use the same production harness for both models."",""prompt_style"":""official Microsoft evaluation prompts"",""sampling"":""pass@1; benchmark-specific where stated"",""temperature"":""source does not state globally"",""tools"":""GitHub Copilot / VS Code production tools""}",False,official_model_card,verified,Displayed exactly as 62.7. Official Microsoft-reported result.
claude-sonnet-4.6,advancedif_rubric_level,86.0,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""maximum sequence length"",""effort"":""maximum reasoning effort"",""harness"":""Microsoft independent evaluation suite"",""judge"":""benchmark-specific Microsoft implementation"",""mode"":""thinking"",""notes"":""Sonnet 4.6 result generated by Microsoft's own evaluation."",""prompt_style"":""benchmark-specific"",""sampling"":""source does not state trial count"",""temperature"":""source does not state"",""tools"":""none; benchmark-provided tools where required""}",False,official_paper,verified,Displayed exactly as 86. Official Microsoft-reported result.
claude-sonnet-4.6,air_bench_2024,88.0,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""maximum sequence length"",""effort"":""maximum reasoning effort"",""harness"":""Microsoft independent evaluation suite"",""judge"":""benchmark-specific Microsoft implementation"",""mode"":""thinking"",""notes"":""Sonnet 4.6 result generated by Microsoft's own evaluation."",""prompt_style"":""benchmark-specific"",""sampling"":""source does not state trial count"",""temperature"":""source does not state"",""tools"":""none; benchmark-provided tools where required""}",False,official_paper,verified,Displayed exactly as 88. Official Microsoft-reported result.
claude-sonnet-4.6,bfcl_v3,76.0,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""maximum sequence length"",""effort"":""maximum reasoning effort"",""harness"":""Microsoft independent evaluation suite"",""judge"":""benchmark-specific Microsoft implementation"",""mode"":""thinking"",""notes"":""Sonnet 4.6 result generated by Microsoft's own evaluation."",""prompt_style"":""benchmark-specific"",""sampling"":""source does not state trial count"",""temperature"":""0.001; top_p=0.97"",""tools"":""none; benchmark-provided tools where required""}",False,official_paper,verified,Displayed exactly as 76. Official Microsoft-reported result.
claude-sonnet-4.6,corpusqa_gpt54_judge,79.0,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""128K maximum output tokens"",""effort"":""maximum reasoning effort"",""harness"":""Microsoft independent evaluation suite"",""judge"":""benchmark-specific Microsoft implementation"",""mode"":""thinking"",""notes"":""Sonnet 4.6 result generated by Microsoft's own evaluation."",""prompt_style"":""benchmark-specific"",""sampling"":""average of 4 independent evaluations"",""temperature"":""source does not state"",""tools"":""none; benchmark-provided tools where required""}",False,official_paper,verified,Displayed exactly as 79. Official Microsoft-reported result.
claude-sonnet-4.6,cyberseceval4_autocomplete,56.0,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""maximum sequence length"",""effort"":""maximum reasoning effort"",""harness"":""Microsoft independent evaluation suite"",""judge"":""benchmark-specific Microsoft implementation"",""mode"":""thinking"",""notes"":""Sonnet 4.6 result generated by Microsoft's own evaluation."",""prompt_style"":""benchmark-specific"",""sampling"":""source does not state trial count"",""temperature"":""source does not state"",""tools"":""none; benchmark-provided tools where required""}",False,official_paper,verified,Displayed exactly as 56. Official Microsoft-reported result.
claude-sonnet-4.6,cyberseceval4_instruct,62.0,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""maximum sequence length"",""effort"":""maximum reasoning effort"",""harness"":""Microsoft independent evaluation suite"",""judge"":""benchmark-specific Microsoft implementation"",""mode"":""thinking"",""notes"":""Sonnet 4.6 result generated by Microsoft's own evaluation."",""prompt_style"":""benchmark-specific"",""sampling"":""source does not state trial count"",""temperature"":""source does not state"",""tools"":""none; benchmark-provided tools where required""}",False,official_paper,verified,Displayed exactly as 62. Official Microsoft-reported result.
claude-sonnet-4.6,graphwalks_lt128k_combined,96.0,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""maximum sequence length"",""effort"":""maximum reasoning effort"",""harness"":""Microsoft independent evaluation suite"",""judge"":""benchmark-specific Microsoft implementation"",""mode"":""thinking"",""notes"":""Sonnet 4.6 result generated by Microsoft's own evaluation."",""prompt_style"":""benchmark-specific"",""sampling"":""source does not state trial count"",""temperature"":""source does not state"",""tools"":""none; benchmark-provided tools where required""}",False,official_paper,verified,Displayed exactly as 96. Official Microsoft-reported result.
claude-sonnet-4.6,longbench_v2_256k,66.0,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""128K maximum output tokens"",""effort"":""maximum reasoning effort"",""harness"":""Microsoft independent evaluation suite"",""judge"":""benchmark-specific Microsoft implementation"",""mode"":""thinking"",""notes"":""Sonnet 4.6 result generated by Microsoft's own evaluation."",""prompt_style"":""benchmark-specific"",""sampling"":""average of 4 independent evaluations"",""temperature"":""source does not state"",""tools"":""none; benchmark-provided tools where required""}",False,official_paper,verified,Displayed exactly as 66. Official Microsoft-reported result.
claude-sonnet-4.6,longfact_claim_precision,98.0,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""maximum sequence length"",""effort"":""maximum reasoning effort"",""harness"":""Microsoft independent evaluation suite"",""judge"":""benchmark-specific Microsoft implementation"",""mode"":""thinking"",""notes"":""Sonnet 4.6 result generated by Microsoft's own evaluation."",""prompt_style"":""benchmark-specific"",""sampling"":""source does not state trial count"",""temperature"":""source does not state"",""tools"":""none; benchmark-provided tools where required""}",False,official_paper,verified,Displayed exactly as 98. Official Microsoft-reported result.
claude-sonnet-4.6,medxpertqa_text,49.0,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""maximum sequence length"",""effort"":""maximum reasoning effort"",""harness"":""Microsoft independent evaluation suite"",""judge"":""benchmark-specific Microsoft implementation"",""mode"":""thinking"",""notes"":""Sonnet 4.6 result generated by Microsoft's own evaluation."",""prompt_style"":""benchmark-specific"",""sampling"":""source does not state trial count"",""temperature"":""source does not state"",""tools"":""none; benchmark-provided tools where required""}",False,official_paper,verified,Displayed exactly as 49. Official Microsoft-reported result.
claude-sonnet-4.6,multichallenge,57.0,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""maximum sequence length"",""effort"":""maximum reasoning effort"",""harness"":""Microsoft independent evaluation suite"",""judge"":""benchmark-specific Microsoft implementation"",""mode"":""thinking"",""notes"":""Sonnet 4.6 result generated by Microsoft's own evaluation."",""prompt_style"":""benchmark-specific"",""sampling"":""source does not state trial count"",""temperature"":""source does not state"",""tools"":""none; benchmark-provided tools where required""}",False,official_paper,verified,Displayed exactly as 57. Official Microsoft-reported result.
claude-sonnet-4.6,simpleqa_verified,29.0,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""maximum sequence length"",""effort"":""maximum reasoning effort"",""harness"":""Microsoft independent evaluation suite"",""judge"":""benchmark-specific Microsoft implementation"",""mode"":""thinking"",""notes"":""Sonnet 4.6 result generated by Microsoft's own evaluation."",""prompt_style"":""benchmark-specific"",""sampling"":""source does not state trial count"",""temperature"":""source does not state"",""tools"":""none; benchmark-provided tools where required""}",False,official_paper,verified,Displayed exactly as 29. Official Microsoft-reported result.
claude-sonnet-4.6,truthfulqa_mc,88.0,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""maximum sequence length"",""effort"":""maximum reasoning effort"",""harness"":""Microsoft independent evaluation suite"",""judge"":""benchmark-specific Microsoft implementation"",""mode"":""thinking"",""notes"":""Sonnet 4.6 result generated by Microsoft's own evaluation."",""prompt_style"":""benchmark-specific"",""sampling"":""source does not state trial count"",""temperature"":""source does not state"",""tools"":""none; benchmark-provided tools where required""}",False,official_paper,verified,Displayed exactly as 88. Official Microsoft-reported result.
mai-code-1-flash,advancedif_average,71.4,https://microsoft.ai/pdf/MAI-Code-1-Flash-Model-Card.PDF,"{""context"":""256K"",""effort"":""adaptive solution length control"",""harness"":""Microsoft MAI-Code-1-Flash comparative evaluation"",""judge"":""benchmark-specific"",""mode"":""adaptive reasoning"",""notes"":""Official instruction-following or agentic-tool-use score; benchmark=advancedif_average."",""prompt_style"":""source does not state"",""sampling"":""single/multi-turn average where applicable"",""temperature"":""source does not state"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly as 71.4. Official Microsoft-reported result.
mai-code-1-flash,aime_2026,92.5,https://microsoft.ai/pdf/MAI-Code-1-Flash-Model-Card.PDF,"{""context"":""256K model context; per-task usage reported separately"",""effort"":""adaptive solution length control"",""harness"":""Microsoft MAI-Code-1-Flash comparative evaluation"",""judge"":""benchmark-specific"",""mode"":""adaptive reasoning"",""notes"":""Accuracy column from the official model-card/release table; benchmark=aime_2026."",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly as 92.5. Official Microsoft-reported result.
mai-code-1-flash,amo_bench,40.0,https://microsoft.ai/pdf/MAI-Code-1-Flash-Model-Card.PDF,"{""context"":""256K model context; per-task usage reported separately"",""effort"":""adaptive solution length control"",""harness"":""Microsoft MAI-Code-1-Flash comparative evaluation"",""judge"":""benchmark-specific"",""mode"":""adaptive reasoning"",""notes"":""Accuracy column from the official model-card/release table; benchmark=amo_bench."",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly as 40.0. Official Microsoft-reported result.
mai-code-1-flash,artifactsbench,36.4,https://microsoft.ai/pdf/MAI-Code-1-Flash-Model-Card.PDF,"{""context"":""256K model context; per-task usage reported separately"",""effort"":""adaptive solution length control"",""harness"":""Microsoft MAI-Code-1-Flash comparative evaluation"",""judge"":""benchmark-specific"",""mode"":""adaptive reasoning"",""notes"":""Accuracy column from the official model-card/release table; benchmark=artifactsbench."",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly as 36.4. Official Microsoft-reported result.
mai-code-1-flash,frontiermath,6.3,https://microsoft.ai/pdf/MAI-Code-1-Flash-Model-Card.PDF,"{""context"":""256K model context; per-task usage reported separately"",""effort"":""adaptive solution length control"",""harness"":""Microsoft MAI-Code-1-Flash comparative evaluation"",""judge"":""benchmark-specific"",""mode"":""adaptive reasoning"",""notes"":""Accuracy column from the official model-card/release table; benchmark=frontiermath."",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly as 6.3. Official Microsoft-reported result.
mai-code-1-flash,gpqa_diamond,84.6,https://microsoft.ai/pdf/MAI-Code-1-Flash-Model-Card.PDF,"{""context"":""256K model context; per-task usage reported separately"",""effort"":""adaptive solution length control"",""harness"":""Microsoft MAI-Code-1-Flash comparative evaluation"",""judge"":""benchmark-specific"",""mode"":""adaptive reasoning"",""notes"":""Accuracy column from the official model-card/release table; benchmark=gpqa_diamond."",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly as 84.6. Official Microsoft-reported result.
mai-code-1-flash,hle_text,18.0,https://microsoft.ai/pdf/MAI-Code-1-Flash-Model-Card.PDF,"{""context"":""256K model context; per-task usage reported separately"",""effort"":""adaptive solution length control"",""harness"":""Microsoft MAI-Code-1-Flash comparative evaluation"",""judge"":""benchmark-specific"",""mode"":""adaptive reasoning"",""notes"":""Accuracy column from the official model-card/release table; benchmark=hle_text."",""prompt_style"":""source does not state"",""sampling"":""source does not state trial count"",""temperature"":""source does not state"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly as 18.0. Official Microsoft-reported result.
mai-code-1-flash,ifbench_single_multiturn_average,75.0,https://microsoft.ai/pdf/MAI-Code-1-Flash-Model-Card.PDF,"{""context"":""256K"",""effort"":""adaptive solution length control"",""harness"":""Microsoft MAI-Code-1-Flash comparative evaluation"",""judge"":""benchmark-specific"",""mode"":""adaptive reasoning"",""notes"":""Official instruction-following or agentic-tool-use score; benchmark=ifbench_single_multiturn_average."",""prompt_style"":""source does not state"",""sampling"":""single/multi-turn average where applicable"",""temperature"":""source does not state"",""tools"":""none""}",False,official_model_card,verified,Displayed exactly as 75.0. Official Microsoft-reported result.
mai-code-1-flash,swe_bench_multilingual,65.5,https://microsoft.ai/pdf/MAI-Code-1-Flash-Model-Card.PDF,"{""context"":""256K"",""effort"":""adaptive solution length control"",""harness"":""GitHub Copilot VS Code production harness"",""judge"":""benchmark-specified"",""mode"":""adaptive reasoning"",""notes"":""Coding-focused 137B-total / 5B-active model. Core coding scores use the same production harness for both models."",""prompt_style"":""official Microsoft evaluation prompts"",""sampling"":""pass@1; benchmark-specific where stated"",""temperature"":""source does not state globally"",""tools"":""GitHub Copilot / VS Code production tools""}",True,official_model_card,verified,Displayed exactly as 65.5. Official Microsoft-reported result.
mai-code-1-flash,swe_bench_pro,51.2,https://microsoft.ai/pdf/MAI-Code-1-Flash-Model-Card.PDF,"{""context"":""256K"",""effort"":""adaptive solution length control"",""harness"":""GitHub Copilot VS Code production harness"",""judge"":""benchmark-specified"",""mode"":""adaptive reasoning"",""notes"":""Coding-focused 137B-total / 5B-active model. Core coding scores use the same production harness for both models."",""prompt_style"":""official Microsoft evaluation prompts"",""sampling"":""pass@1; benchmark-specific where stated"",""temperature"":""source does not state globally"",""tools"":""GitHub Copilot / VS Code production tools""}",True,official_model_card,verified,Displayed exactly as 51.2. Official Microsoft-reported result.
mai-code-1-flash,swe_bench_verified,71.6,https://microsoft.ai/pdf/MAI-Code-1-Flash-Model-Card.PDF,"{""context"":""256K"",""effort"":""adaptive solution length control"",""harness"":""GitHub Copilot VS Code production harness"",""judge"":""benchmark-specified"",""mode"":""adaptive reasoning"",""notes"":""Coding-focused 137B-total / 5B-active model. Core coding scores use the same production harness for both models."",""prompt_style"":""official Microsoft evaluation prompts"",""sampling"":""pass@1; benchmark-specific where stated"",""temperature"":""source does not state globally"",""tools"":""GitHub Copilot / VS Code production tools""}",True,official_model_card,verified,Displayed exactly as 71.6. Official Microsoft-reported result.
mai-code-1-flash,tau2_bench_telecom,71.7,https://microsoft.ai/pdf/MAI-Code-1-Flash-Model-Card.PDF,"{""context"":""256K"",""effort"":""adaptive solution length control"",""harness"":""Microsoft MAI-Code-1-Flash comparative evaluation"",""judge"":""benchmark-specific"",""mode"":""adaptive reasoning"",""notes"":""Official instruction-following or agentic-tool-use score; benchmark=tau2_bench_telecom."",""prompt_style"":""source does not state"",""sampling"":""single/multi-turn average where applicable"",""temperature"":""source does not state"",""tools"":""Tau2 telecom tools""}",False,official_model_card,verified,Displayed exactly as 71.7. Official Microsoft-reported result.
mai-code-1-flash,terminal_bench,54.8,https://microsoft.ai/pdf/MAI-Code-1-Flash-Model-Card.PDF,"{""context"":""256K"",""effort"":""adaptive solution length control"",""harness"":""GitHub Copilot VS Code production harness"",""judge"":""benchmark-specified"",""mode"":""adaptive reasoning"",""notes"":""Coding-focused 137B-total / 5B-active model. Core coding scores use the same production harness for both models."",""prompt_style"":""official Microsoft evaluation prompts"",""sampling"":""pass@1; benchmark-specific where stated"",""temperature"":""source does not state globally"",""tools"":""GitHub Copilot / VS Code production tools""}",True,official_model_card,verified,Displayed exactly as 54.8. Official Microsoft-reported result.
mai-thinking-1,advancedif_rubric_level,85.0,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""maximum sequence length"",""effort"":""maximum reasoning effort"",""harness"":""Microsoft official evaluation suite"",""judge"":""calibrated LLM judge"",""mode"":""thinking"",""notes"":""AdvancedIF rubric-level score, not all-rubrics pass rate."",""prompt_style"":""benchmark-specific Microsoft implementation"",""sampling"":""average of 4 runs"",""temperature"":""1.0; top_p=0.97"",""tools"":""none; benchmark-provided tools for BFCL v3""}",False,official_paper,verified,Displayed exactly as 85. Official Microsoft-reported result.
mai-thinking-1,aime_2025,97.0,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""256K maximum output tokens"",""effort"":""maximum available reasoning"",""harness"":""Microsoft STEM evaluation"",""judge"":""regex/SymPy with GPT-5-mini fallback"",""mode"":""thinking"",""notes"":""Technical report headline evaluation configuration."",""prompt_style"":""boxed final answer for STEM; ReAct for agentic coding"",""sampling"":""average of 4 runs; temperature=1; top_p=0.97"",""temperature"":""1.0; top_p=0.97"",""tools"":""none""}",False,official_paper,verified,Displayed exactly as 97.0. Official Microsoft-reported result.
mai-thinking-1,aime_2026,94.5,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""256K maximum output tokens"",""effort"":""maximum available reasoning"",""harness"":""Microsoft STEM evaluation"",""judge"":""regex/SymPy with GPT-5-mini fallback"",""mode"":""thinking"",""notes"":""Technical report headline evaluation configuration."",""prompt_style"":""boxed final answer for STEM; ReAct for agentic coding"",""sampling"":""average of 4 runs; temperature=1; top_p=0.97"",""temperature"":""1.0; top_p=0.97"",""tools"":""none""}",False,official_paper,verified,Displayed exactly as 94.5. Official Microsoft-reported result.
mai-thinking-1,air_bench_2024,88.0,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""maximum sequence length"",""effort"":""maximum reasoning effort"",""harness"":""Microsoft official evaluation suite"",""judge"":""benchmark-specific"",""mode"":""thinking"",""notes"":""Table 12 configuration from the technical report."",""prompt_style"":""benchmark-specific Microsoft implementation"",""sampling"":""average of 4 runs"",""temperature"":""1.0; top_p=0.97"",""tools"":""none; benchmark-provided tools for BFCL v3""}",False,official_paper,verified,Displayed exactly as 88. Official Microsoft-reported result.
mai-thinking-1,bfcl_v3,72.0,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""maximum sequence length"",""effort"":""maximum reasoning effort"",""harness"":""Microsoft official evaluation suite"",""judge"":""benchmark-specific"",""mode"":""thinking"",""notes"":""Table 12 configuration from the technical report."",""prompt_style"":""benchmark-specific Microsoft implementation"",""sampling"":""average of 4 runs"",""temperature"":""0.001; top_p=0.97"",""tools"":""BFCL-provided functions""}",False,official_paper,verified,Displayed exactly as 72. Official Microsoft-reported result.
mai-thinking-1,corpusqa_gpt54_judge,82.0,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""128K maximum output; input limit not stated"",""effort"":""maximum reasoning effort"",""harness"":""Microsoft long-context evaluation"",""judge"":""GPT-5.4 high"",""mode"":""thinking"",""notes"":""Technical report Table 19 long-context evaluation."",""prompt_style"":""source does not state"",""sampling"":""average of 4 independent evaluations"",""temperature"":""1.0; top_p=0.97"",""tools"":""none""}",False,official_paper,verified,Displayed exactly as 82. Official Microsoft-reported result.
mai-thinking-1,cyberseceval4_autocomplete,63.0,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""maximum sequence length"",""effort"":""maximum reasoning effort"",""harness"":""Microsoft official evaluation suite"",""judge"":""benchmark-specific"",""mode"":""thinking"",""notes"":""Table 12 configuration from the technical report."",""prompt_style"":""benchmark-specific Microsoft implementation"",""sampling"":""average of 4 runs"",""temperature"":""1.0; top_p=0.97"",""tools"":""none; benchmark-provided tools for BFCL v3""}",False,official_paper,verified,Displayed exactly as 63. Official Microsoft-reported result.
mai-thinking-1,cyberseceval4_instruct,63.0,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""maximum sequence length"",""effort"":""maximum reasoning effort"",""harness"":""Microsoft official evaluation suite"",""judge"":""benchmark-specific"",""mode"":""thinking"",""notes"":""Table 12 configuration from the technical report."",""prompt_style"":""benchmark-specific Microsoft implementation"",""sampling"":""average of 4 runs"",""temperature"":""1.0; top_p=0.97"",""tools"":""none; benchmark-provided tools for BFCL v3""}",False,official_paper,verified,Displayed exactly as 63. Official Microsoft-reported result.
mai-thinking-1,gpqa_diamond,84.2,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""256K maximum output tokens"",""effort"":""maximum available reasoning"",""harness"":""Microsoft STEM evaluation"",""judge"":""regex/SymPy with GPT-5-mini fallback"",""mode"":""thinking"",""notes"":""Technical report headline evaluation configuration."",""prompt_style"":""boxed final answer for STEM; ReAct for agentic coding"",""sampling"":""pass@1 averaged over 16 rollouts; temperature=1; top_p=0.97"",""temperature"":""1.0; top_p=0.97"",""tools"":""none""}",False,official_paper,verified,Displayed exactly as 84.2. Official Microsoft-reported result.
mai-thinking-1,graphwalks_lt128k_combined,90.0,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""<=128K measured with o200k_base tokenizer"",""effort"":""maximum reasoning effort"",""harness"":""Microsoft official evaluation suite"",""judge"":""answer-set F1"",""mode"":""thinking"",""notes"":""Table 12 configuration from the technical report."",""prompt_style"":""benchmark-specific Microsoft implementation"",""sampling"":""average of 4 runs"",""temperature"":""1.0; top_p=0.97"",""tools"":""none; benchmark-provided tools for BFCL v3""}",False,official_paper,verified,Displayed exactly as 90. Official Microsoft-reported result.
mai-thinking-1,healthbench_professional_length_adjusted,35.0,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""maximum sequence length"",""effort"":""maximum reasoning effort"",""harness"":""Microsoft official evaluation suite"",""judge"":""GPT-5.4 grader with official rubrics"",""mode"":""thinking"",""notes"":""Length-penalized HealthBench Professional primary metric."",""prompt_style"":""benchmark-specific Microsoft implementation"",""sampling"":""average of 4 runs"",""temperature"":""1.0; top_p=0.97"",""tools"":""none; benchmark-provided tools for BFCL v3""}",False,official_paper,verified,Displayed exactly as 35. Official Microsoft-reported result.
mai-thinking-1,hmmt_feb_2026,84.9,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""256K maximum output tokens"",""effort"":""maximum available reasoning"",""harness"":""Microsoft STEM evaluation"",""judge"":""regex/SymPy with GPT-5-mini fallback"",""mode"":""thinking"",""notes"":""Technical report headline evaluation configuration."",""prompt_style"":""boxed final answer for STEM; ReAct for agentic coding"",""sampling"":""average of 4 runs; temperature=1; top_p=0.97"",""temperature"":""1.0; top_p=0.97"",""tools"":""none""}",False,official_paper,verified,Displayed exactly as 84.9. Official Microsoft-reported result.
mai-thinking-1,ifbench,69.0,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""maximum sequence length"",""effort"":""maximum reasoning effort"",""harness"":""Microsoft official evaluation suite"",""judge"":""benchmark-specific"",""mode"":""thinking"",""notes"":""Table 12 configuration from the technical report."",""prompt_style"":""benchmark-specific Microsoft implementation"",""sampling"":""average of 4 runs"",""temperature"":""1.0; top_p=0.97"",""tools"":""none; benchmark-provided tools for BFCL v3""}",False,official_paper,verified,Displayed exactly as 69. Official Microsoft-reported result.
mai-thinking-1,livecodebench_v6,87.7,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""256K maximum output tokens"",""effort"":""maximum available reasoning"",""harness"":""Microsoft STEM evaluation"",""judge"":""regex/SymPy with GPT-5-mini fallback"",""mode"":""thinking"",""notes"":""Technical report headline evaluation configuration."",""prompt_style"":""boxed final answer for STEM; ReAct for agentic coding"",""sampling"":""average of 4 runs; temperature=1; top_p=0.97"",""temperature"":""1.0; top_p=0.97"",""tools"":""none""}",False,official_paper,verified,Displayed exactly as 87.7. Official Microsoft-reported result.
mai-thinking-1,longbench_v2_256k,61.0,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""256K maximum input; 128K maximum output"",""effort"":""maximum reasoning effort"",""harness"":""Microsoft long-context evaluation"",""judge"":""exact multiple-choice answer matching"",""mode"":""thinking"",""notes"":""Technical report Table 19 long-context evaluation."",""prompt_style"":""source does not state"",""sampling"":""average of 4 independent evaluations"",""temperature"":""1.0; top_p=0.97"",""tools"":""none""}",False,official_paper,verified,Displayed exactly as 61. Official Microsoft-reported result.
mai-thinking-1,longfact_claim_precision,98.0,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""maximum sequence length"",""effort"":""maximum reasoning effort"",""harness"":""Microsoft official evaluation suite"",""judge"":""simplified claim extraction plus LLM judge"",""mode"":""thinking"",""notes"":""Table 12 configuration from the technical report."",""prompt_style"":""benchmark-specific Microsoft implementation"",""sampling"":""average of 4 runs"",""temperature"":""1.0; top_p=0.97"",""tools"":""none; benchmark-provided tools for BFCL v3""}",False,official_paper,verified,Displayed exactly as 98. Official Microsoft-reported result.
mai-thinking-1,medxpertqa_text,43.0,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""maximum sequence length"",""effort"":""maximum reasoning effort"",""harness"":""Microsoft official evaluation suite"",""judge"":""GPT-5.4 parses the answer letter"",""mode"":""thinking"",""notes"":""2,450 expert-curated 10-choice text questions."",""prompt_style"":""benchmark-specific Microsoft implementation"",""sampling"":""average of 4 runs"",""temperature"":""1.0; top_p=0.97"",""tools"":""none; benchmark-provided tools for BFCL v3""}",False,official_paper,verified,Displayed exactly as 43. Official Microsoft-reported result.
mai-thinking-1,mmlu_pro,85.0,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""maximum sequence length"",""effort"":""maximum reasoning effort"",""harness"":""Microsoft official evaluation suite"",""judge"":""benchmark-specific"",""mode"":""thinking"",""notes"":""Table 12 configuration from the technical report."",""prompt_style"":""benchmark-specific Microsoft implementation"",""sampling"":""average of 4 runs"",""temperature"":""1.0; top_p=0.97"",""tools"":""none; benchmark-provided tools for BFCL v3""}",False,official_paper,verified,Displayed exactly as 85. Official Microsoft-reported result.
mai-thinking-1,multichallenge,53.0,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""maximum sequence length"",""effort"":""maximum reasoning effort"",""harness"":""Microsoft official evaluation suite"",""judge"":""benchmark-specific"",""mode"":""thinking"",""notes"":""Table 12 configuration from the technical report."",""prompt_style"":""benchmark-specific Microsoft implementation"",""sampling"":""average of 4 runs"",""temperature"":""1.0; top_p=0.97"",""tools"":""none; benchmark-provided tools for BFCL v3""}",False,official_paper,verified,Displayed exactly as 53. Official Microsoft-reported result.
mai-thinking-1,simpleqa_verified,31.0,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""maximum sequence length"",""effort"":""maximum reasoning effort"",""harness"":""Microsoft official evaluation suite"",""judge"":""benchmark-specific"",""mode"":""thinking"",""notes"":""Table 12 configuration from the technical report."",""prompt_style"":""benchmark-specific Microsoft implementation"",""sampling"":""average of 4 runs"",""temperature"":""1.0; top_p=0.97"",""tools"":""none; benchmark-provided tools for BFCL v3""}",False,official_paper,verified,Displayed exactly as 31. Official Microsoft-reported result.
mai-thinking-1,swe_bench_pro,52.8,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""256K total context; 8K max output; 1,000 max steps"",""effort"":""maximum available reasoning"",""harness"":""Microsoft ReAct agent evaluation"",""judge"":""benchmark verifier"",""mode"":""thinking"",""notes"":""Technical report headline evaluation configuration."",""prompt_style"":""boxed final answer for STEM; ReAct for agentic coding"",""sampling"":""average of 4 runs; temperature=1; top_p=0.97"",""temperature"":""1.0; top_p=0.97"",""tools"":""ReAct loop with bash and string-replacement tools""}",False,official_paper,verified,Displayed exactly as 52.8. Official Microsoft-reported result.
mai-thinking-1,swe_bench_verified,73.5,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""256K total context; 8K max output; 1,000 max steps"",""effort"":""maximum available reasoning"",""harness"":""Microsoft ReAct agent evaluation"",""judge"":""benchmark verifier"",""mode"":""thinking"",""notes"":""Technical report headline evaluation configuration."",""prompt_style"":""boxed final answer for STEM; ReAct for agentic coding"",""sampling"":""average of 4 runs; temperature=1; top_p=0.97"",""temperature"":""1.0; top_p=0.97"",""tools"":""ReAct loop with bash and string-replacement tools""}",False,official_paper,verified,Displayed exactly as 73.5. Official Microsoft-reported result.
mai-thinking-1,terminal_bench,46.0,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""256K total context; 32K max output; 1,000 max steps"",""effort"":""maximum available reasoning"",""harness"":""Microsoft ReAct agent evaluation"",""judge"":""benchmark verifier"",""mode"":""thinking"",""notes"":""Technical report headline evaluation configuration."",""prompt_style"":""boxed final answer for STEM; ReAct for agentic coding"",""sampling"":""average of 4 runs; temperature=1; top_p=0.97"",""temperature"":""1.0; top_p=0.97"",""tools"":""ReAct loop with bash only""}",False,official_paper,verified,Displayed exactly as 46.0. Official Microsoft-reported result.
mai-thinking-1,truthfulqa_mc,88.0,https://microsoft.ai/wp-content/uploads/2026/06/main_20260602_2.pdf,"{""context"":""maximum sequence length"",""effort"":""maximum reasoning effort"",""harness"":""Microsoft official evaluation suite"",""judge"":""benchmark-specific"",""mode"":""thinking"",""notes"":""Table 12 configuration from the technical report."",""prompt_style"":""benchmark-specific Microsoft implementation"",""sampling"":""average of 4 runs"",""temperature"":""1.0; top_p=0.97"",""tools"":""none; benchmark-provided tools for BFCL v3""}",False,official_paper,verified,Displayed exactly as 88. Official Microsoft-reported result.
claude-opus-4.6,claw_eval_multimodal_pass3,24.8,https://raw.githubusercontent.com/claw-eval/claw-eval.github.io/19d378a646d892ca6c8148691aea02da1b184eb8/assets/benchmark-0fIVfCD7.js,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official Claw-Eval v1.1 leaderboard"",""input_modalities"":""image and text"",""judge"":""full-trajectory benchmark grading"",""mode"":""source does not state"",""prompt_style"":""official Claw-Eval task prompts"",""sampling"":""Pass^3; all three trajectories must pass"",""task_count"":101,""task_subset"":""multimodal 101-task subset"",""temperature"":""source does not state"",""tools"":""official Claw-Eval agent environment"",""trials"":""3 per task; 303 generations""}",False,official_model_card,verified,Exact official leaderboard field for the multimodal 101-task subset. Model mode/effort is not inferred.
claude-opus-4.6,claw_eval_pass3,70.4,https://raw.githubusercontent.com/claw-eval/claw-eval.github.io/19d378a646d892ca6c8148691aea02da1b184eb8/assets/benchmark-0fIVfCD7.js,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official Claw-Eval v1.1 leaderboard"",""input_modalities"":""text"",""judge"":""full-trajectory benchmark grading"",""mode"":""source does not state"",""prompt_style"":""official Claw-Eval task prompts"",""sampling"":""Pass^3; all three trajectories must pass"",""task_count"":199,""task_subset"":""non-multimodal 199-task subset"",""temperature"":""source does not state"",""tools"":""official Claw-Eval agent environment"",""trials"":""3 per task; 597 generations""}",False,official_blog,verified,Exact official leaderboard field for the non-multimodal 199-task subset. Model mode/effort is not inferred.
claude-sonnet-4.6,claw_eval_multimodal_pass3,23.8,https://raw.githubusercontent.com/claw-eval/claw-eval.github.io/19d378a646d892ca6c8148691aea02da1b184eb8/assets/benchmark-0fIVfCD7.js,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official Claw-Eval v1.1 leaderboard"",""input_modalities"":""image and text"",""judge"":""full-trajectory benchmark grading"",""mode"":""source does not state"",""prompt_style"":""official Claw-Eval task prompts"",""sampling"":""Pass^3; all three trajectories must pass"",""task_count"":101,""task_subset"":""multimodal 101-task subset"",""temperature"":""source does not state"",""tools"":""official Claw-Eval agent environment"",""trials"":""3 per task; 303 generations""}",False,official_model_card,verified,Exact official leaderboard field for the multimodal 101-task subset. Model mode/effort is not inferred.
claude-sonnet-4.6,claw_eval_pass3,68.3,https://raw.githubusercontent.com/claw-eval/claw-eval.github.io/19d378a646d892ca6c8148691aea02da1b184eb8/assets/benchmark-0fIVfCD7.js,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official Claw-Eval v1.1 leaderboard"",""input_modalities"":""text"",""judge"":""full-trajectory benchmark grading"",""mode"":""source does not state"",""prompt_style"":""official Claw-Eval task prompts"",""sampling"":""Pass^3; all three trajectories must pass"",""task_count"":161,""task_subset"":""core-general 161-task subset"",""temperature"":""source does not state"",""tools"":""official Claw-Eval agent environment"",""trials"":""3 per task; 483 generations""}",False,official_blog,verified,"Average tokens per trajectory=142233 is a physical efficiency annotation, not a benchmark score or score setting. Exact official leaderboard field for the core-general 161-task subset. Model mode/effort is not inferred."
deepseek-v3.2,claw_eval_pass3,42.2,https://raw.githubusercontent.com/claw-eval/claw-eval.github.io/19d378a646d892ca6c8148691aea02da1b184eb8/assets/benchmark-0fIVfCD7.js,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official Claw-Eval v1.1 leaderboard"",""input_modalities"":""text"",""judge"":""full-trajectory benchmark grading"",""mode"":""source does not state"",""prompt_style"":""official Claw-Eval task prompts"",""sampling"":""Pass^3; all three trajectories must pass"",""task_count"":161,""task_subset"":""core-general 161-task subset"",""temperature"":""source does not state"",""tools"":""official Claw-Eval agent environment"",""trials"":""3 per task; 483 generations""}",False,official_blog,verified,"Average tokens per trajectory=237262 is a physical efficiency annotation, not a benchmark score or score setting. Exact official leaderboard field for the core-general 161-task subset. Model mode/effort is not inferred."
deepseek-v4-flash,bigbench_hard,86.9,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""source does not state"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""3-shot few-shot continuation"",""sampling"":""3-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The Xiaomi table does not state this comparator Base checkpoint's context length, and no comparator Base config is locked; no Pro context value or conflict is copied onto this observation."
deepseek-v4-flash,c_eval,92.1,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""source does not state"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""5-shot few-shot continuation"",""sampling"":""5-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The Xiaomi table does not state this comparator Base checkpoint's context length, and no comparator Base config is locked; no Pro context value or conflict is copied onto this observation."
deepseek-v4-flash,cmmlu,90.4,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""source does not state"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""5-shot few-shot continuation"",""sampling"":""5-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The Xiaomi table does not state this comparator Base checkpoint's context length, and no comparator Base config is locked; no Pro context value or conflict is copied onto this observation."
deepseek-v4-flash,drop,88.6,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""source does not state"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""3-shot few-shot continuation"",""sampling"":""3-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The Xiaomi table does not state this comparator Base checkpoint's context length, and no comparator Base config is locked; no Pro context value or conflict is copied onto this observation."
deepseek-v4-flash,gsm8k,90.8,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""source does not state"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""8-shot few-shot continuation"",""sampling"":""8-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The Xiaomi table does not state this comparator Base checkpoint's context length, and no comparator Base config is locked; no Pro context value or conflict is copied onto this observation."
deepseek-v4-flash,hellaswag,85.7,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""source does not state"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""10-shot few-shot continuation"",""sampling"":""10-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The Xiaomi table does not state this comparator Base checkpoint's context length, and no comparator Base config is locked; no Pro context value or conflict is copied onto this observation."
deepseek-v4-flash,math,57.4,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""source does not state"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""4-shot few-shot continuation"",""sampling"":""4-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The Xiaomi table does not state this comparator Base checkpoint's context length, and no comparator Base config is locked; no Pro context value or conflict is copied onto this observation."
deepseek-v4-flash,mmlu,88.7,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""source does not state"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""5-shot few-shot continuation"",""sampling"":""5-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The Xiaomi table does not state this comparator Base checkpoint's context length, and no comparator Base config is locked; no Pro context value or conflict is copied onto this observation."
deepseek-v4-flash,mmlu_redux,89.4,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""source does not state"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""5-shot few-shot continuation"",""sampling"":""5-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The Xiaomi table does not state this comparator Base checkpoint's context length, and no comparator Base config is locked; no Pro context value or conflict is copied onto this observation."
deepseek-v4-pro,bigbench_hard,87.5,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""source does not state"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""3-shot few-shot continuation"",""sampling"":""3-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The Xiaomi table does not state this comparator Base checkpoint's context length, and no comparator Base config is locked; no Pro context value or conflict is copied onto this observation."
deepseek-v4-pro,c_eval,93.1,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""source does not state"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""5-shot few-shot continuation"",""sampling"":""5-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The Xiaomi table does not state this comparator Base checkpoint's context length, and no comparator Base config is locked; no Pro context value or conflict is copied onto this observation."
deepseek-v4-pro,cmmlu,90.8,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""source does not state"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""5-shot few-shot continuation"",""sampling"":""5-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The Xiaomi table does not state this comparator Base checkpoint's context length, and no comparator Base config is locked; no Pro context value or conflict is copied onto this observation."
deepseek-v4-pro,drop,88.7,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""source does not state"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""3-shot few-shot continuation"",""sampling"":""3-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The Xiaomi table does not state this comparator Base checkpoint's context length, and no comparator Base config is locked; no Pro context value or conflict is copied onto this observation."
deepseek-v4-pro,gsm8k,92.6,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""source does not state"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""8-shot few-shot continuation"",""sampling"":""8-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The Xiaomi table does not state this comparator Base checkpoint's context length, and no comparator Base config is locked; no Pro context value or conflict is copied onto this observation."
deepseek-v4-pro,hellaswag,88.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""source does not state"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""10-shot few-shot continuation"",""sampling"":""10-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The Xiaomi table does not state this comparator Base checkpoint's context length, and no comparator Base config is locked; no Pro context value or conflict is copied onto this observation."
deepseek-v4-pro,hle,37.7,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro/resolve/b5968e9190ef611bbf34a7229255be88a0e937c1/README.md,"{""context"":""source does not state"",""effort"":""Max (DeepSeek-V4-Pro-Max)"",""harness"":""source does not state"",""input_modalities"":""source does not state"",""judge"":""source does not state"",""mode"":""thinking (Think Max)"",""prompt_style"":""source does not state"",""sampling"":""Pass@1"",""temperature"":""source does not state"",""tools"":""source does not state"",""trials"":""source does not state""}",False,official_blog,verified,Semantic alternative printed inside the same physical HLE cell; it does not add a physical position. Exact DeepSeek official model-card score from the DeepSeek-V4-Pro-Max table. Only the source-backed Think Max mode and Max effort are retained; benchmark-specific harness and unstated sampling details remain unknown.
deepseek-v4-pro,hle_tools,48.2,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro/resolve/b5968e9190ef611bbf34a7229255be88a0e937c1/README.md,"{""context"":""source does not state"",""effort"":""Max (DeepSeek-V4-Pro-Max)"",""harness"":""source does not state"",""input_modalities"":""source does not state"",""judge"":""source does not state"",""mode"":""thinking (Think Max)"",""prompt_style"":""source does not state"",""sampling"":""Pass@1"",""temperature"":""source does not state"",""tools"":""with tools; exact tool suite source does not state"",""trials"":""source does not state""}",False,official_blog,verified,DeepSeek V4 Pro is explicitly reported at max effort. Exact DeepSeek official model-card score from the DeepSeek-V4-Pro-Max table. Only the source-backed Think Max mode and Max effort are retained; benchmark-specific harness and unstated sampling details remain unknown.
deepseek-v4-pro,math,64.5,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""source does not state"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""4-shot few-shot continuation"",""sampling"":""4-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The Xiaomi table does not state this comparator Base checkpoint's context length, and no comparator Base config is locked; no Pro context value or conflict is copied onto this observation."
deepseek-v4-pro,mmlu,90.1,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""source does not state"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""5-shot few-shot continuation"",""sampling"":""5-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The Xiaomi table does not state this comparator Base checkpoint's context length, and no comparator Base config is locked; no Pro context value or conflict is copied onto this observation."
deepseek-v4-pro,mmlu_redux,90.8,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""source does not state"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""5-shot few-shot continuation"",""sampling"":""5-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The Xiaomi table does not state this comparator Base checkpoint's context length, and no comparator Base config is locked; no Pro context value or conflict is copied onto this observation."
gemini-3-flash,claw_eval_pass3,48.4,https://raw.githubusercontent.com/claw-eval/claw-eval.github.io/19d378a646d892ca6c8148691aea02da1b184eb8/assets/benchmark-0fIVfCD7.js,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official Claw-Eval v1.1 leaderboard"",""input_modalities"":""text"",""judge"":""full-trajectory benchmark grading"",""mode"":""source does not state"",""prompt_style"":""official Claw-Eval task prompts"",""sampling"":""Pass^3; all three trajectories must pass"",""task_count"":161,""task_subset"":""core-general 161-task subset"",""temperature"":""source does not state"",""tools"":""official Claw-Eval agent environment"",""trials"":""3 per task; 483 generations""}",False,official_blog,verified,"Average tokens per trajectory=82126 is a physical efficiency annotation, not a benchmark score or score setting. Exact official leaderboard field for the core-general 161-task subset. Model mode/effort is not inferred."
glm-5.1,hle,31.0,https://z.ai/blog/assets/glm-5.1-B3CLmgrT.js,"{""context"":""source/provider setting"",""effort"":""official reported setting"",""harness"":""official Z.ai GLM-5.1 evaluation"",""input_modalities"":""text"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""provider official evaluation prompt"",""sampling"":""pass@1"",""temperature"":""source/provider setting"",""tools"":""none"",""trials"":""source-reported""}",False,official_blog,verified,Semantic alternative printed inside the same physical HLE cell; it does not add a physical position. Exact provider-official score and reported setting.
glm-5.1,hle_tools,52.3,https://z.ai/blog/assets/glm-5.1-B3CLmgrT.js,"{""context"":""source/provider setting"",""effort"":""official reported setting"",""harness"":""official Z.ai GLM-5.1 evaluation"",""input_modalities"":""text"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""provider official evaluation prompt"",""sampling"":""pass@1"",""temperature"":""source/provider setting"",""tools"":""source-reported search/code/web tools"",""trials"":""source-reported""}",False,official_blog,verified,Exact provider-official score and reported setting.
gpt-5.4,graphwalks_bfs_bucket_0_4k,100.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""0\u20134K"",""context_window_variant"":""standard"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""n=58""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
gpt-5.4,graphwalks_bfs_bucket_128k_256k,57.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""128K\u2013256K"",""context_window_variant"":""standard"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""n=100""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
gpt-5.4,graphwalks_bfs_bucket_16k_32k,92.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""16K\u201332K"",""context_window_variant"":""standard"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""n=49""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
gpt-5.4,graphwalks_bfs_bucket_32k_64k,90.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""32K\u201364K"",""context_window_variant"":""standard"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""n=50""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
gpt-5.4,graphwalks_bfs_bucket_4k_8k,100.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""4K\u20138K"",""context_window_variant"":""standard"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""n=77""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
gpt-5.4,graphwalks_bfs_bucket_64k_128k,78.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""64K\u2013128K"",""context_window_variant"":""standard"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""n=51""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
gpt-5.4,graphwalks_bfs_bucket_8k_16k,100.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""8K\u201316K"",""context_window_variant"":""standard"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""n=84""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
gpt-5.4,graphwalks_parents_bucket_0_4k,100.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""0\u20134K"",""context_window_variant"":""standard"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""n=120""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
gpt-5.4,graphwalks_parents_bucket_128k_256k,85.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""128K\u2013256K"",""context_window_variant"":""standard"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""n=100""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
gpt-5.4,graphwalks_parents_bucket_16k_32k,100.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""16K\u201332K"",""context_window_variant"":""standard"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""n=50""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
gpt-5.4,graphwalks_parents_bucket_32k_64k,99.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""32K\u201364K"",""context_window_variant"":""standard"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""n=50""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
gpt-5.4,graphwalks_parents_bucket_4k_8k,100.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""4K\u20138K"",""context_window_variant"":""standard"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""n=76""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
gpt-5.4,graphwalks_parents_bucket_64k_128k,89.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""64K\u2013128K"",""context_window_variant"":""standard"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""n=50""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
gpt-5.4,graphwalks_parents_bucket_8k_16k,100.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""8K\u201316K"",""context_window_variant"":""standard"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""n=54""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
kimi-k2,bigbench_hard,88.7,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""source does not state"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""3-shot few-shot continuation"",""sampling"":""3-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The Xiaomi table does not state this comparator Base checkpoint's context length, and no comparator Base config is locked; no Pro context value or conflict is copied onto this observation."
kimi-k2,c_eval,92.5,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""source does not state"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""5-shot few-shot continuation"",""sampling"":""5-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The Xiaomi table does not state this comparator Base checkpoint's context length, and no comparator Base config is locked; no Pro context value or conflict is copied onto this observation."
kimi-k2,cmmlu,90.9,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""source does not state"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""5-shot few-shot continuation"",""sampling"":""5-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The Xiaomi table does not state this comparator Base checkpoint's context length, and no comparator Base config is locked; no Pro context value or conflict is copied onto this observation."
kimi-k2,drop,83.6,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""source does not state"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""3-shot few-shot continuation"",""sampling"":""3-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The Xiaomi table does not state this comparator Base checkpoint's context length, and no comparator Base config is locked; no Pro context value or conflict is copied onto this observation."
kimi-k2,humaneval_plus,84.8,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""source does not state"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""1-shot few-shot continuation"",""sampling"":""1-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The Xiaomi table does not state this comparator Base checkpoint's context length, and no comparator Base config is locked; no Pro context value or conflict is copied onto this observation."
kimi-k2,mbpp_plus,73.8,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""source does not state"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""3-shot few-shot continuation"",""sampling"":""3-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The Xiaomi table does not state this comparator Base checkpoint's context length, and no comparator Base config is locked; no Pro context value or conflict is copied onto this observation."
kimi-k2.5,claw_eval_pass3,52.8,https://raw.githubusercontent.com/claw-eval/claw-eval.github.io/19d378a646d892ca6c8148691aea02da1b184eb8/assets/benchmark-0fIVfCD7.js,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official Claw-Eval v1.1 leaderboard"",""input_modalities"":""text"",""judge"":""full-trajectory benchmark grading"",""mode"":""source does not state"",""prompt_style"":""official Claw-Eval task prompts"",""sampling"":""Pass^3; all three trajectories must pass"",""task_count"":161,""task_subset"":""core-general 161-task subset"",""temperature"":""source does not state"",""tools"":""official Claw-Eval agent environment"",""trials"":""3 per task; 483 generations""}",False,official_blog,verified,"Average tokens per trajectory=97512 is a physical efficiency annotation, not a benchmark score or score setting. Exact official leaderboard field for the core-general 161-task subset. Model mode/effort is not inferred."
kimi-k2.6,hle_tools,54.0,https://huggingface.co/moonshotai/Kimi-K2.6/resolve/d9cb81bc88b9bd2dc89877599f23c614ded72b9c/README.md,"{""context"":""262,144 (256K)"",""effort"":""max; up to 98,304 generation tokens"",""harness"":""official Kimi K2.6 evaluation"",""input_modalities"":""text"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""provider official evaluation prompt"",""sampling"":""pass@1"",""temperature"":""1.0; top_p=1.0"",""tools"":""source-reported search/code/web tools"",""trials"":""source-reported""}",False,official_blog,verified,Exact provider-official score and reported setting.
mimo-v2-omni,claw_eval_multimodal_pass3,15.8,https://raw.githubusercontent.com/claw-eval/claw-eval.github.io/19d378a646d892ca6c8148691aea02da1b184eb8/assets/benchmark-0fIVfCD7.js,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official Claw-Eval v1.1 leaderboard"",""input_modalities"":""image and text"",""judge"":""full-trajectory benchmark grading"",""mode"":""source does not state"",""prompt_style"":""official Claw-Eval task prompts"",""sampling"":""Pass^3; all three trajectories must pass"",""task_count"":101,""task_subset"":""multimodal 101-task subset"",""temperature"":""source does not state"",""tools"":""official Claw-Eval agent environment"",""trials"":""3 per task; 303 generations""}",False,official_model_card,verified,Exact official leaderboard field for the multimodal 101-task subset. Model mode/effort is not inferred.
mimo-v2-omni,claw_eval_pass3,52.2,https://raw.githubusercontent.com/claw-eval/claw-eval.github.io/19d378a646d892ca6c8148691aea02da1b184eb8/assets/benchmark-0fIVfCD7.js,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official Claw-Eval v1.1 leaderboard"",""input_modalities"":""text"",""judge"":""full-trajectory benchmark grading"",""mode"":""source does not state"",""prompt_style"":""official Claw-Eval task prompts"",""sampling"":""Pass^3; all three trajectories must pass"",""task_count"":161,""task_subset"":""core-general 161-task subset"",""temperature"":""source does not state"",""tools"":""official Claw-Eval agent environment"",""trials"":""3 per task; 483 generations""}",False,official_blog,verified,"Average tokens per trajectory=141781 is a physical efficiency annotation, not a benchmark score or score setting. Exact official leaderboard field for the core-general 161-task subset. Model mode/effort is not inferred."
mimo-v2-omni,graphwalks_bfs_128k,68.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5/resolve/63651580ca774f8504f676040460aed3e1244ac1/assets/mimo-v2.5-graphwalks.jpeg,"{""context"":""128K"",""context_window_variant"":""exact-length line chart"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo-V2.5 GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.68; normalized to 68.0 percent without changing meaning. Source prints 0.68 on [0,1]; campaign stores the equivalent 68.0%."
mimo-v2-omni,graphwalks_bfs_256k,54.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5/resolve/63651580ca774f8504f676040460aed3e1244ac1/assets/mimo-v2.5-graphwalks.jpeg,"{""context"":""256K"",""context_window_variant"":""exact-length line chart"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo-V2.5 GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.54; normalized to 54.0 percent without changing meaning. Source prints 0.54 on [0,1]; campaign stores the equivalent 54.0%."
mimo-v2-omni,graphwalks_bfs_32k,87.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5/resolve/63651580ca774f8504f676040460aed3e1244ac1/assets/mimo-v2.5-graphwalks.jpeg,"{""context"":""32K"",""context_window_variant"":""exact-length line chart"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo-V2.5 GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.87; normalized to 87.0 percent without changing meaning. Source prints 0.87 on [0,1]; campaign stores the equivalent 87.0%."
mimo-v2-omni,graphwalks_bfs_64k,79.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5/resolve/63651580ca774f8504f676040460aed3e1244ac1/assets/mimo-v2.5-graphwalks.jpeg,"{""context"":""64K"",""context_window_variant"":""exact-length line chart"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo-V2.5 GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.79; normalized to 79.0 percent without changing meaning. Source prints 0.79 on [0,1]; campaign stores the equivalent 79.0%."
mimo-v2-omni,graphwalks_parents_128k,56.99999999999999,https://huggingface.co/XiaomiMiMo/MiMo-V2.5/resolve/63651580ca774f8504f676040460aed3e1244ac1/assets/mimo-v2.5-graphwalks.jpeg,"{""context"":""128K"",""context_window_variant"":""exact-length line chart"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo-V2.5 GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.57; normalized to 57.0 percent without changing meaning. Source prints 0.57 on [0,1]; campaign stores the equivalent 57.0%."
mimo-v2-omni,graphwalks_parents_256k,17.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5/resolve/63651580ca774f8504f676040460aed3e1244ac1/assets/mimo-v2.5-graphwalks.jpeg,"{""context"":""256K"",""context_window_variant"":""exact-length line chart"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo-V2.5 GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.17; normalized to 17.0 percent without changing meaning. Source prints 0.17 on [0,1]; campaign stores the equivalent 17.0%."
mimo-v2-omni,graphwalks_parents_32k,85.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5/resolve/63651580ca774f8504f676040460aed3e1244ac1/assets/mimo-v2.5-graphwalks.jpeg,"{""context"":""32K"",""context_window_variant"":""exact-length line chart"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo-V2.5 GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.85; normalized to 85.0 percent without changing meaning. Source prints 0.85 on [0,1]; campaign stores the equivalent 85.0%."
mimo-v2-omni,graphwalks_parents_64k,70.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5/resolve/63651580ca774f8504f676040460aed3e1244ac1/assets/mimo-v2.5-graphwalks.jpeg,"{""context"":""64K"",""context_window_variant"":""exact-length line chart"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo-V2.5 GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.70; normalized to 70.0 percent without changing meaning. Source prints 0.70 on [0,1]; campaign stores the equivalent 70.0%."
mimo-v2-pro,claw_eval_pass3,57.8,https://raw.githubusercontent.com/claw-eval/claw-eval.github.io/19d378a646d892ca6c8148691aea02da1b184eb8/assets/benchmark-0fIVfCD7.js,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official Claw-Eval v1.1 leaderboard"",""input_modalities"":""text"",""judge"":""full-trajectory benchmark grading"",""mode"":""source does not state"",""prompt_style"":""official Claw-Eval task prompts"",""sampling"":""Pass^3; all three trajectories must pass"",""task_count"":199,""task_subset"":""non-multimodal 199-task subset"",""temperature"":""source does not state"",""tools"":""official Claw-Eval agent environment"",""trials"":""3 per task; 597 generations""}",False,official_blog,verified,Exact official leaderboard field for the non-multimodal 199-task subset. Model mode/effort is not inferred.
mimo-v2-pro,graphwalks_bfs_128k,68.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/assets/post_training_evaluation.jpg,"{""context"":""128K"",""context_window_variant"":""exact-length line chart"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo-V2.5-Pro GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.68; normalized to 68.0 percent without changing meaning. Source prints 0.68 on [0,1]; campaign stores the equivalent 68.0%."
mimo-v2-pro,graphwalks_bfs_1m,0.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/assets/post_training_evaluation.jpg,"{""context"":""1M"",""context_window_variant"":""exact-length line chart"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo-V2.5-Pro GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.00; normalized to 0.0 percent without changing meaning. Source prints 0.00 on [0,1]; campaign stores the equivalent 0.0%."
mimo-v2-pro,graphwalks_bfs_256k,44.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/assets/post_training_evaluation.jpg,"{""context"":""256K"",""context_window_variant"":""exact-length line chart"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo-V2.5-Pro GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.44; normalized to 44.0 percent without changing meaning. Source prints 0.44 on [0,1]; campaign stores the equivalent 44.0%."
mimo-v2-pro,graphwalks_bfs_32k,81.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/assets/post_training_evaluation.jpg,"{""context"":""32K"",""context_window_variant"":""exact-length line chart"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo-V2.5-Pro GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.81; normalized to 81.0 percent without changing meaning. Source prints 0.81 on [0,1]; campaign stores the equivalent 81.0%."
mimo-v2-pro,graphwalks_bfs_512k,12.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/assets/post_training_evaluation.jpg,"{""context"":""512K"",""context_window_variant"":""exact-length line chart"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo-V2.5-Pro GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.12; normalized to 12.0 percent without changing meaning. Source prints 0.12 on [0,1]; campaign stores the equivalent 12.0%."
mimo-v2-pro,graphwalks_bfs_64k,75.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/assets/post_training_evaluation.jpg,"{""context"":""64K"",""context_window_variant"":""exact-length line chart"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo-V2.5-Pro GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.75; normalized to 75.0 percent without changing meaning. Source prints 0.75 on [0,1]; campaign stores the equivalent 75.0%."
mimo-v2-pro,graphwalks_bfs_bucket_0_4k,100.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""0\u20134K"",""context_window_variant"":""standard"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""n=50""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
mimo-v2-pro,graphwalks_bfs_bucket_128k_256k,46.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""128K\u2013256K"",""context_window_variant"":""standard"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""n=50""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
mimo-v2-pro,graphwalks_bfs_bucket_16k_32k,66.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""16K\u201332K"",""context_window_variant"":""standard"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""n=50""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
mimo-v2-pro,graphwalks_bfs_bucket_32k_64k,62.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""32K\u201364K"",""context_window_variant"":""standard"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""n=50""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
mimo-v2-pro,graphwalks_bfs_bucket_4k_8k,98.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""4K\u20138K"",""context_window_variant"":""standard"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""n=50""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
mimo-v2-pro,graphwalks_bfs_bucket_64k_128k,56.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""64K\u2013128K"",""context_window_variant"":""standard"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""n=50""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
mimo-v2-pro,graphwalks_bfs_bucket_8k_16k,76.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""8K\u201316K"",""context_window_variant"":""standard"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""n=50""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
mimo-v2-pro,graphwalks_parents_128k,44.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/assets/post_training_evaluation.jpg,"{""context"":""128K"",""context_window_variant"":""exact-length line chart"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo-V2.5-Pro GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.44; normalized to 44.0 percent without changing meaning. Source prints 0.44 on [0,1]; campaign stores the equivalent 44.0%."
mimo-v2-pro,graphwalks_parents_1m,0.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/assets/post_training_evaluation.jpg,"{""context"":""1M"",""context_window_variant"":""exact-length line chart"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo-V2.5-Pro GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.00; normalized to 0.0 percent without changing meaning. Source prints 0.00 on [0,1]; campaign stores the equivalent 0.0%."
mimo-v2-pro,graphwalks_parents_256k,35.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/assets/post_training_evaluation.jpg,"{""context"":""256K"",""context_window_variant"":""exact-length line chart"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo-V2.5-Pro GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.35; normalized to 35.0 percent without changing meaning. Source prints 0.35 on [0,1]; campaign stores the equivalent 35.0%."
mimo-v2-pro,graphwalks_parents_32k,79.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/assets/post_training_evaluation.jpg,"{""context"":""32K"",""context_window_variant"":""exact-length line chart"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo-V2.5-Pro GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.79; normalized to 79.0 percent without changing meaning. Source prints 0.79 on [0,1]; campaign stores the equivalent 79.0%."
mimo-v2-pro,graphwalks_parents_512k,14.000000000000002,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/assets/post_training_evaluation.jpg,"{""context"":""512K"",""context_window_variant"":""exact-length line chart"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo-V2.5-Pro GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.14; normalized to 14.0 percent without changing meaning. Source prints 0.14 on [0,1]; campaign stores the equivalent 14.0%."
mimo-v2-pro,graphwalks_parents_64k,63.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/assets/post_training_evaluation.jpg,"{""context"":""64K"",""context_window_variant"":""exact-length line chart"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo-V2.5-Pro GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.63; normalized to 63.0 percent without changing meaning. Source prints 0.63 on [0,1]; campaign stores the equivalent 63.0%."
mimo-v2-pro,graphwalks_parents_bucket_0_4k,100.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""0\u20134K"",""context_window_variant"":""standard"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""n=100""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
mimo-v2-pro,graphwalks_parents_bucket_128k_256k,14.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""128K\u2013256K"",""context_window_variant"":""standard"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""n=50""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
mimo-v2-pro,graphwalks_parents_bucket_16k_32k,80.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""16K\u201332K"",""context_window_variant"":""standard"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""n=50""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
mimo-v2-pro,graphwalks_parents_bucket_32k_64k,56.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""32K\u201364K"",""context_window_variant"":""standard"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""n=50""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
mimo-v2-pro,graphwalks_parents_bucket_4k_8k,100.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""4K\u20138K"",""context_window_variant"":""standard"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""n=100""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
mimo-v2-pro,graphwalks_parents_bucket_64k_128k,47.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""64K\u2013128K"",""context_window_variant"":""standard"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""n=50""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
mimo-v2-pro,graphwalks_parents_bucket_8k_16k,98.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""8K\u201316K"",""context_window_variant"":""standard"",""effort"":""source does not state"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""source does not state"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""n=100""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
mimo-v2.5,arc_challenge,96.5,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""262,144 (256K)"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""25-shot few-shot continuation"",""sampling"":""25-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The locked MiMo-V2.5-Base config exposes max_position_embeddings=262,144, consistent with the 256K download/table description; there is no context conflict."
mimo-v2.5,bigbench_hard,87.2,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""262,144 (256K)"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""3-shot few-shot continuation"",""sampling"":""3-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The locked MiMo-V2.5-Base config exposes max_position_embeddings=262,144, consistent with the 256K download/table description; there is no context conflict."
mimo-v2.5,c_eval,88.6,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""262,144 (256K)"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""5-shot few-shot continuation"",""sampling"":""5-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The locked MiMo-V2.5-Base config exposes max_position_embeddings=262,144, consistent with the 256K download/table description; there is no context conflict."
mimo-v2.5,charxiv_reasoning,81.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5/resolve/63651580ca774f8504f676040460aed3e1244ac1/assets/mimo-v2.5-multimodal-bench.png,"{""context"":""1,048,576"",""effort"":""default"",""harness"":""official Xiaomi MiMo family benchmark evaluation"",""input_modalities"":""image and text"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""Xiaomi MiMo official evaluation prompt"",""sampling"":""pass@1; source does not state repeated-trial count"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,Locked Xiaomi observation for CharXiv RQ.
mimo-v2.5,claw_eval_multimodal_pass3,23.8,https://raw.githubusercontent.com/claw-eval/claw-eval.github.io/19d378a646d892ca6c8148691aea02da1b184eb8/assets/benchmark-0fIVfCD7.js,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official Claw-Eval v1.1 leaderboard"",""input_modalities"":""image and text"",""judge"":""full-trajectory benchmark grading"",""mode"":""source does not state"",""prompt_style"":""official Claw-Eval task prompts"",""sampling"":""Pass^3; all three trajectories must pass"",""task_count"":101,""task_subset"":""multimodal 101-task subset"",""temperature"":""source does not state"",""tools"":""official Claw-Eval agent environment"",""trials"":""3 per task; 303 generations""}",False,official_model_card,verified,Exact official leaderboard field for the multimodal 101-task subset. Model mode/effort is not inferred.
mimo-v2.5,claw_eval_pass3,62.3,https://raw.githubusercontent.com/claw-eval/claw-eval.github.io/19d378a646d892ca6c8148691aea02da1b184eb8/assets/benchmark-0fIVfCD7.js,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official Claw-Eval v1.1 leaderboard"",""input_modalities"":""text"",""judge"":""full-trajectory benchmark grading"",""mode"":""source does not state"",""prompt_style"":""official Claw-Eval task prompts"",""sampling"":""Pass^3; all three trajectories must pass"",""task_count"":199,""task_subset"":""non-multimodal 199-task subset"",""temperature"":""source does not state"",""tools"":""official Claw-Eval agent environment"",""trials"":""3 per task; 597 generations""}",False,official_blog,verified,Exact official leaderboard field for the non-multimodal 199-task subset. Model mode/effort is not inferred.
mimo-v2.5,cmmlu,88.2,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""262,144 (256K)"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""5-shot few-shot continuation"",""sampling"":""5-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The locked MiMo-V2.5-Base config exposes max_position_embeddings=262,144, consistent with the 256K download/table description; there is no context conflict."
mimo-v2.5,drop,83.7,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""262,144 (256K)"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""3-shot few-shot continuation"",""sampling"":""3-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The locked MiMo-V2.5-Base config exposes max_position_embeddings=262,144, consistent with the 256K download/table description; there is no context conflict."
mimo-v2.5,gpqa_diamond,58.1,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""262,144 (256K)"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""5-shot few-shot continuation"",""sampling"":""5-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The locked MiMo-V2.5-Base config exposes max_position_embeddings=262,144, consistent with the 256K download/table description; there is no context conflict."
mimo-v2.5,graphwalks_bfs_128k,75.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5/resolve/63651580ca774f8504f676040460aed3e1244ac1/assets/mimo-v2.5-graphwalks.jpeg,"{""context"":""128K"",""context_window_variant"":""exact-length line chart"",""effort"":""default"",""harness"":""Xiaomi MiMo-V2.5 GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.75; normalized to 75.0 percent without changing meaning. Source prints 0.75 on [0,1]; campaign stores the equivalent 75.0%."
mimo-v2.5,graphwalks_bfs_1m,54.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5/resolve/63651580ca774f8504f676040460aed3e1244ac1/assets/mimo-v2.5-graphwalks.jpeg,"{""context"":""1M"",""context_window_variant"":""exact-length line chart"",""effort"":""default"",""harness"":""Xiaomi MiMo-V2.5 GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.54; normalized to 54.0 percent without changing meaning. Source prints 0.54 on [0,1]; campaign stores the equivalent 54.0%."
mimo-v2.5,graphwalks_bfs_256k,78.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5/resolve/63651580ca774f8504f676040460aed3e1244ac1/assets/mimo-v2.5-graphwalks.jpeg,"{""context"":""256K"",""context_window_variant"":""exact-length line chart"",""effort"":""default"",""harness"":""Xiaomi MiMo-V2.5 GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.78; normalized to 78.0 percent without changing meaning. Source prints 0.78 on [0,1]; campaign stores the equivalent 78.0%."
mimo-v2.5,graphwalks_bfs_32k,86.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5/resolve/63651580ca774f8504f676040460aed3e1244ac1/assets/mimo-v2.5-graphwalks.jpeg,"{""context"":""32K"",""context_window_variant"":""exact-length line chart"",""effort"":""default"",""harness"":""Xiaomi MiMo-V2.5 GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.86; normalized to 86.0 percent without changing meaning. Source prints 0.86 on [0,1]; campaign stores the equivalent 86.0%."
mimo-v2.5,graphwalks_bfs_512k,57.99999999999999,https://huggingface.co/XiaomiMiMo/MiMo-V2.5/resolve/63651580ca774f8504f676040460aed3e1244ac1/assets/mimo-v2.5-graphwalks.jpeg,"{""context"":""512K"",""context_window_variant"":""exact-length line chart"",""effort"":""default"",""harness"":""Xiaomi MiMo-V2.5 GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.58; normalized to 58.0 percent without changing meaning. Source prints 0.58 on [0,1]; campaign stores the equivalent 58.0%."
mimo-v2.5,graphwalks_bfs_64k,85.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5/resolve/63651580ca774f8504f676040460aed3e1244ac1/assets/mimo-v2.5-graphwalks.jpeg,"{""context"":""64K"",""context_window_variant"":""exact-length line chart"",""effort"":""default"",""harness"":""Xiaomi MiMo-V2.5 GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.85; normalized to 85.0 percent without changing meaning. Source prints 0.85 on [0,1]; campaign stores the equivalent 85.0%."
mimo-v2.5,graphwalks_parents_128k,100.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5/resolve/63651580ca774f8504f676040460aed3e1244ac1/assets/mimo-v2.5-graphwalks.jpeg,"{""context"":""128K"",""context_window_variant"":""exact-length line chart"",""effort"":""default"",""harness"":""Xiaomi MiMo-V2.5 GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 1.00; normalized to 100.0 percent without changing meaning. Source prints 1.00 on [0,1]; campaign stores the equivalent 100.0%."
mimo-v2.5,graphwalks_parents_1m,87.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5/resolve/63651580ca774f8504f676040460aed3e1244ac1/assets/mimo-v2.5-graphwalks.jpeg,"{""context"":""1M"",""context_window_variant"":""exact-length line chart"",""effort"":""default"",""harness"":""Xiaomi MiMo-V2.5 GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.87; normalized to 87.0 percent without changing meaning. Source prints 0.87 on [0,1]; campaign stores the equivalent 87.0%."
mimo-v2.5,graphwalks_parents_256k,98.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5/resolve/63651580ca774f8504f676040460aed3e1244ac1/assets/mimo-v2.5-graphwalks.jpeg,"{""context"":""256K"",""context_window_variant"":""exact-length line chart"",""effort"":""default"",""harness"":""Xiaomi MiMo-V2.5 GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.98; normalized to 98.0 percent without changing meaning. Source prints 0.98 on [0,1]; campaign stores the equivalent 98.0%."
mimo-v2.5,graphwalks_parents_32k,100.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5/resolve/63651580ca774f8504f676040460aed3e1244ac1/assets/mimo-v2.5-graphwalks.jpeg,"{""context"":""32K"",""context_window_variant"":""exact-length line chart"",""effort"":""default"",""harness"":""Xiaomi MiMo-V2.5 GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 1.00; normalized to 100.0 percent without changing meaning. Source prints 1.00 on [0,1]; campaign stores the equivalent 100.0%."
mimo-v2.5,graphwalks_parents_512k,97.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5/resolve/63651580ca774f8504f676040460aed3e1244ac1/assets/mimo-v2.5-graphwalks.jpeg,"{""context"":""512K"",""context_window_variant"":""exact-length line chart"",""effort"":""default"",""harness"":""Xiaomi MiMo-V2.5 GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.97; normalized to 97.0 percent without changing meaning. Source prints 0.97 on [0,1]; campaign stores the equivalent 97.0%."
mimo-v2.5,graphwalks_parents_64k,100.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5/resolve/63651580ca774f8504f676040460aed3e1244ac1/assets/mimo-v2.5-graphwalks.jpeg,"{""context"":""64K"",""context_window_variant"":""exact-length line chart"",""effort"":""default"",""harness"":""Xiaomi MiMo-V2.5 GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 1.00; normalized to 100.0 percent without changing meaning. Source prints 1.00 on [0,1]; campaign stores the equivalent 100.0%."
mimo-v2.5,gsm8k,83.3,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""262,144 (256K)"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""8-shot few-shot continuation"",""sampling"":""8-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The locked MiMo-V2.5-Base config exposes max_position_embeddings=262,144, consistent with the 256K download/table description; there is no context conflict."
mimo-v2.5,hellaswag,88.6,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""262,144 (256K)"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""10-shot few-shot continuation"",""sampling"":""10-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The locked MiMo-V2.5-Base config exposes max_position_embeddings=262,144, consistent with the 256K download/table description; there is no context conflict."
mimo-v2.5,humaneval_plus,71.3,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""262,144 (256K)"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""1-shot few-shot continuation"",""sampling"":""1-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The locked MiMo-V2.5-Base config exposes max_position_embeddings=262,144, consistent with the 256K download/table description; there is no context conflict."
mimo-v2.5,livecodebench_v6,35.5,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""262,144 (256K)"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""1-shot few-shot continuation"",""sampling"":""1-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The locked MiMo-V2.5-Base config exposes max_position_embeddings=262,144, consistent with the 256K download/table description; there is no context conflict."
mimo-v2.5,math,67.7,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""262,144 (256K)"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""4-shot few-shot continuation"",""sampling"":""4-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The locked MiMo-V2.5-Base config exposes max_position_embeddings=262,144, consistent with the 256K download/table description; there is no context conflict."
mimo-v2.5,mbpp_plus,70.9,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""262,144 (256K)"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""3-shot few-shot continuation"",""sampling"":""3-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The locked MiMo-V2.5-Base config exposes max_position_embeddings=262,144, consistent with the 256K download/table description; there is no context conflict."
mimo-v2.5,mmlu,86.3,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""262,144 (256K)"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""5-shot few-shot continuation"",""sampling"":""5-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The locked MiMo-V2.5-Base config exposes max_position_embeddings=262,144, consistent with the 256K download/table description; there is no context conflict."
mimo-v2.5,mmlu_pro,65.8,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""262,144 (256K)"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""5-shot few-shot continuation"",""sampling"":""5-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The locked MiMo-V2.5-Base config exposes max_position_embeddings=262,144, consistent with the 256K download/table description; there is no context conflict."
mimo-v2.5,mmlu_redux,89.8,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""262,144 (256K)"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""5-shot few-shot continuation"",""sampling"":""5-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The locked MiMo-V2.5-Base config exposes max_position_embeddings=262,144, consistent with the 256K download/table description; there is no context conflict."
mimo-v2.5,mmmu_pro,77.9,https://huggingface.co/XiaomiMiMo/MiMo-V2.5/resolve/63651580ca774f8504f676040460aed3e1244ac1/assets/mimo-v2.5-multimodal-bench.png,"{""context"":""1,048,576"",""effort"":""default"",""harness"":""official Xiaomi MiMo family benchmark evaluation"",""input_modalities"":""image and text"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""Xiaomi MiMo official evaluation prompt"",""sampling"":""pass@1; source does not state repeated-trial count"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,Locked Xiaomi observation for MMMU-Pro.
mimo-v2.5,swe_bench_pro,56.1,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""1,048,576"",""effort"":""default"",""harness"":""official Xiaomi MiMo family benchmark evaluation"",""input_modalities"":""text"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""Xiaomi MiMo official evaluation prompt"",""sampling"":""source does not state repeated-trial count"",""temperature"":""1.0; top_p=0.95"",""tools"":""agentic code/terminal tools"",""trials"":""source does not state""}",False,official_blog,verified,Locked Xiaomi observation for SWE-Bench Pro.
mimo-v2.5,tau3_bench,69.5,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""1,048,576"",""effort"":""default"",""harness"":""official Xiaomi MiMo family benchmark evaluation"",""input_modalities"":""text"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""Xiaomi MiMo official evaluation prompt"",""sampling"":""pass@1; source does not state repeated-trial count"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""source does not state""}",False,official_blog,verified,Locked Xiaomi observation for τ³-bench.
mimo-v2.5,terminal_bench,65.8,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""1,048,576"",""effort"":""default"",""harness"":""official Xiaomi MiMo family benchmark evaluation"",""input_modalities"":""text"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""Xiaomi MiMo official evaluation prompt"",""sampling"":""source does not state repeated-trial count"",""temperature"":""1.0; top_p=0.95"",""tools"":""agentic code/terminal tools"",""trials"":""source does not state""}",False,official_blog,verified,Locked Xiaomi observation for Terminal-Bench 2.0.
mimo-v2.5,video_mme,87.7,https://huggingface.co/XiaomiMiMo/MiMo-V2.5/resolve/63651580ca774f8504f676040460aed3e1244ac1/assets/mimo-v2.5-multimodal-bench.png,"{""context"":""1,048,576"",""effort"":""default"",""harness"":""official Xiaomi MiMo family benchmark evaluation"",""input_modalities"":""video, audio, image, and text"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""Xiaomi MiMo official evaluation prompt"",""sampling"":""pass@1; source does not state repeated-trial count"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,Locked Xiaomi observation for Video-MME.
mimo-v2.5,videoholmes,64.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5/resolve/63651580ca774f8504f676040460aed3e1244ac1/assets/mimo-v2.5-multimodal-bench.png,"{""context"":""1,048,576"",""effort"":""default"",""harness"":""official Xiaomi MiMo family benchmark evaluation"",""input_modalities"":""video, audio, image, and text"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""Xiaomi MiMo official evaluation prompt"",""sampling"":""pass@1; source does not state repeated-trial count"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,Locked Xiaomi observation for VideoHolmes.
mimo-v2.5-pro,arc_challenge,97.2,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""256K (reported score-table/model-download setting)"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""25-shot few-shot continuation"",""sampling"":""25-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The locked MiMo-V2.5-Pro-Base config exposes max_position_embeddings=1,048,576 while the score-table/download surface reports 256K; both exact values and the conflict are preserved."
mimo-v2.5-pro,bigbench_hard,88.4,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""256K (reported score-table/model-download setting)"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""3-shot few-shot continuation"",""sampling"":""3-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The locked MiMo-V2.5-Pro-Base config exposes max_position_embeddings=1,048,576 while the score-table/download surface reports 256K; both exact values and the conflict are preserved."
mimo-v2.5-pro,c_eval,91.5,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""256K (reported score-table/model-download setting)"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""5-shot few-shot continuation"",""sampling"":""5-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The locked MiMo-V2.5-Pro-Base config exposes max_position_embeddings=1,048,576 while the score-table/download surface reports 256K; both exact values and the conflict are preserved."
mimo-v2.5-pro,claw_eval_pass3,63.8,https://raw.githubusercontent.com/claw-eval/claw-eval.github.io/19d378a646d892ca6c8148691aea02da1b184eb8/assets/benchmark-0fIVfCD7.js,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official Claw-Eval v1.1 leaderboard"",""input_modalities"":""text"",""judge"":""full-trajectory benchmark grading"",""mode"":""source does not state"",""prompt_style"":""official Claw-Eval task prompts"",""sampling"":""Pass^3; all three trajectories must pass"",""task_count"":199,""task_subset"":""non-multimodal 199-task subset"",""temperature"":""source does not state"",""tools"":""official Claw-Eval agent environment"",""trials"":""3 per task; 597 generations""}",False,official_blog,verified,Exact official leaderboard field for the non-multimodal 199-task subset. Model mode/effort is not inferred.
mimo-v2.5-pro,cmmlu,90.2,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""256K (reported score-table/model-download setting)"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""5-shot few-shot continuation"",""sampling"":""5-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The locked MiMo-V2.5-Pro-Base config exposes max_position_embeddings=1,048,576 while the score-table/download surface reports 256K; both exact values and the conflict are preserved."
mimo-v2.5-pro,drop,86.3,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""256K (reported score-table/model-download setting)"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""3-shot few-shot continuation"",""sampling"":""3-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The locked MiMo-V2.5-Pro-Base config exposes max_position_embeddings=1,048,576 while the score-table/download surface reports 256K; both exact values and the conflict are preserved."
mimo-v2.5-pro,gdpval_aa_elo,1581.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""1,048,576"",""effort"":""default"",""harness"":""official Xiaomi MiMo family benchmark evaluation"",""input_modalities"":""text"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""Xiaomi MiMo official evaluation prompt"",""sampling"":""pass@1; source does not state repeated-trial count"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""source does not state""}",False,official_blog,verified,Locked Xiaomi observation for GDPVal-AA (Elo).
mimo-v2.5-pro,gpqa_diamond,66.7,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""256K (reported score-table/model-download setting)"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""5-shot few-shot continuation"",""sampling"":""5-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The locked MiMo-V2.5-Pro-Base config exposes max_position_embeddings=1,048,576 while the score-table/download surface reports 256K; both exact values and the conflict are preserved."
mimo-v2.5-pro,graphwalks_bfs_128k,82.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/assets/post_training_evaluation.jpg,"{""context"":""128K"",""context_window_variant"":""exact-length line chart"",""effort"":""default"",""harness"":""Xiaomi MiMo-V2.5-Pro GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.82; normalized to 82.0 percent without changing meaning. Source prints 0.82 on [0,1]; campaign stores the equivalent 82.0%."
mimo-v2.5-pro,graphwalks_bfs_1m,37.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/assets/post_training_evaluation.jpg,"{""context"":""1M"",""context_window_variant"":""exact-length line chart"",""effort"":""default"",""harness"":""Xiaomi MiMo-V2.5-Pro GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.37; normalized to 37.0 percent without changing meaning. Source prints 0.37 on [0,1]; campaign stores the equivalent 37.0%."
mimo-v2.5-pro,graphwalks_bfs_256k,77.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/assets/post_training_evaluation.jpg,"{""context"":""256K"",""context_window_variant"":""exact-length line chart"",""effort"":""default"",""harness"":""Xiaomi MiMo-V2.5-Pro GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.77; normalized to 77.0 percent without changing meaning. Source prints 0.77 on [0,1]; campaign stores the equivalent 77.0%."
mimo-v2.5-pro,graphwalks_bfs_32k,81.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/assets/post_training_evaluation.jpg,"{""context"":""32K"",""context_window_variant"":""exact-length line chart"",""effort"":""default"",""harness"":""Xiaomi MiMo-V2.5-Pro GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.81; normalized to 81.0 percent without changing meaning. Source prints 0.81 on [0,1]; campaign stores the equivalent 81.0%."
mimo-v2.5-pro,graphwalks_bfs_512k,56.00000000000001,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/assets/post_training_evaluation.jpg,"{""context"":""512K"",""context_window_variant"":""exact-length line chart"",""effort"":""default"",""harness"":""Xiaomi MiMo-V2.5-Pro GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.56; normalized to 56.0 percent without changing meaning. Source prints 0.56 on [0,1]; campaign stores the equivalent 56.0%."
mimo-v2.5-pro,graphwalks_bfs_64k,79.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/assets/post_training_evaluation.jpg,"{""context"":""64K"",""context_window_variant"":""exact-length line chart"",""effort"":""default"",""harness"":""Xiaomi MiMo-V2.5-Pro GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.79; normalized to 79.0 percent without changing meaning. Source prints 0.79 on [0,1]; campaign stores the equivalent 79.0%."
mimo-v2.5-pro,graphwalks_bfs_bucket_0_4k,94.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""0\u20134K"",""context_window_variant"":""standard"",""effort"":""default"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""n=50""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
mimo-v2.5-pro,graphwalks_bfs_bucket_128k_256k,79.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""128K\u2013256K"",""context_window_variant"":""standard"",""effort"":""default"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""n=50""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
mimo-v2.5-pro,graphwalks_bfs_bucket_16k_32k,89.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""16K\u201332K"",""context_window_variant"":""standard"",""effort"":""default"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""n=50""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
mimo-v2.5-pro,graphwalks_bfs_bucket_256k_512k,62.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""256K\u2013512K"",""context_window_variant"":""1M"",""effort"":""default"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""n=8""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
mimo-v2.5-pro,graphwalks_bfs_bucket_32k_64k,81.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""32K\u201364K"",""context_window_variant"":""standard"",""effort"":""default"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""n=50""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
mimo-v2.5-pro,graphwalks_bfs_bucket_4k_8k,91.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""4K\u20138K"",""context_window_variant"":""standard"",""effort"":""default"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""n=50""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
mimo-v2.5-pro,graphwalks_bfs_bucket_64k_128k,77.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""64K\u2013128K"",""context_window_variant"":""standard"",""effort"":""default"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""n=50""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
mimo-v2.5-pro,graphwalks_bfs_bucket_8k_16k,79.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""8K\u201316K"",""context_window_variant"":""standard"",""effort"":""default"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""n=50""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
mimo-v2.5-pro,graphwalks_parents_128k,99.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/assets/post_training_evaluation.jpg,"{""context"":""128K"",""context_window_variant"":""exact-length line chart"",""effort"":""default"",""harness"":""Xiaomi MiMo-V2.5-Pro GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.99; normalized to 99.0 percent without changing meaning. Source prints 0.99 on [0,1]; campaign stores the equivalent 99.0%."
mimo-v2.5-pro,graphwalks_parents_1m,62.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/assets/post_training_evaluation.jpg,"{""context"":""1M"",""context_window_variant"":""exact-length line chart"",""effort"":""default"",""harness"":""Xiaomi MiMo-V2.5-Pro GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.62; normalized to 62.0 percent without changing meaning. Source prints 0.62 on [0,1]; campaign stores the equivalent 62.0%."
mimo-v2.5-pro,graphwalks_parents_256k,97.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/assets/post_training_evaluation.jpg,"{""context"":""256K"",""context_window_variant"":""exact-length line chart"",""effort"":""default"",""harness"":""Xiaomi MiMo-V2.5-Pro GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.97; normalized to 97.0 percent without changing meaning. Source prints 0.97 on [0,1]; campaign stores the equivalent 97.0%."
mimo-v2.5-pro,graphwalks_parents_32k,98.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/assets/post_training_evaluation.jpg,"{""context"":""32K"",""context_window_variant"":""exact-length line chart"",""effort"":""default"",""harness"":""Xiaomi MiMo-V2.5-Pro GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.98; normalized to 98.0 percent without changing meaning. Source prints 0.98 on [0,1]; campaign stores the equivalent 98.0%."
mimo-v2.5-pro,graphwalks_parents_512k,92.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/assets/post_training_evaluation.jpg,"{""context"":""512K"",""context_window_variant"":""exact-length line chart"",""effort"":""default"",""harness"":""Xiaomi MiMo-V2.5-Pro GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.92; normalized to 92.0 percent without changing meaning. Source prints 0.92 on [0,1]; campaign stores the equivalent 92.0%."
mimo-v2.5-pro,graphwalks_parents_64k,98.0,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/assets/post_training_evaluation.jpg,"{""context"":""64K"",""context_window_variant"":""exact-length line chart"",""effort"":""default"",""harness"":""Xiaomi MiMo-V2.5-Pro GraphWalks exact-length evaluation"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""source does not state""}",False,official_model_card,verified,"Literal chart label 0.98; normalized to 98.0 percent without changing meaning. Source prints 0.98 on [0,1]; campaign stores the equivalent 98.0%."
mimo-v2.5-pro,graphwalks_parents_bucket_0_4k,99.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""0\u20134K"",""context_window_variant"":""standard"",""effort"":""default"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""n=100""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
mimo-v2.5-pro,graphwalks_parents_bucket_128k_256k,95.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""128K\u2013256K"",""context_window_variant"":""standard"",""effort"":""default"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""n=50""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
mimo-v2.5-pro,graphwalks_parents_bucket_16k_32k,100.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""16K\u201332K"",""context_window_variant"":""standard"",""effort"":""default"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""n=50""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
mimo-v2.5-pro,graphwalks_parents_bucket_256k_512k,97.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""256K\u2013512K"",""context_window_variant"":""1M"",""effort"":""default"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""n=3""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
mimo-v2.5-pro,graphwalks_parents_bucket_32k_64k,100.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""32K\u201364K"",""context_window_variant"":""standard"",""effort"":""default"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""n=50""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
mimo-v2.5-pro,graphwalks_parents_bucket_4k_8k,100.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""4K\u20138K"",""context_window_variant"":""standard"",""effort"":""default"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""n=50""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
mimo-v2.5-pro,graphwalks_parents_bucket_64k_128k,99.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""64K\u2013128K"",""context_window_variant"":""standard"",""effort"":""default"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""n=50""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
mimo-v2.5-pro,graphwalks_parents_bucket_8k_16k,100.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""8K\u201316K"",""context_window_variant"":""standard"",""effort"":""default"",""harness"":""Xiaomi MiMo GraphWalks evaluation with Anthropic fixes"",""input_modalities"":""text"",""judge"":""deterministic set-overlap F1 against answer node list"",""mode"":""thinking"",""prompt_style"":""3-shot GraphWalks prompt"",""sampling"":""one response per graph prompt"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""n=50""}",False,official_blog,verified,Exact heatmap bucket. Counts and the 1M variant are preserved; no aggregation into campaign range rows.
mimo-v2.5-pro,gsm8k,99.6,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""256K (reported score-table/model-download setting)"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""8-shot few-shot continuation"",""sampling"":""8-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The locked MiMo-V2.5-Pro-Base config exposes max_position_embeddings=1,048,576 while the score-table/download surface reports 256K; both exact values and the conflict are preserved."
mimo-v2.5-pro,hellaswag,89.8,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""256K (reported score-table/model-download setting)"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""10-shot few-shot continuation"",""sampling"":""10-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The locked MiMo-V2.5-Pro-Base config exposes max_position_embeddings=1,048,576 while the score-table/download surface reports 256K; both exact values and the conflict are preserved."
mimo-v2.5-pro,hle,34.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""1,048,576"",""effort"":""default"",""harness"":""official Xiaomi MiMo family benchmark evaluation"",""input_modalities"":""text and image"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""Xiaomi MiMo official evaluation prompt"",""sampling"":""pass@1; source does not state trial count"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""source does not state""}",False,official_blog,verified,Semantic alternative printed inside the same physical HLE cell; it does not add a physical position. Locked Xiaomi observation for Humanity's Last Exam.
mimo-v2.5-pro,hle_tools,48.0,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""1,048,576"",""effort"":""default"",""harness"":""official Xiaomi MiMo family benchmark evaluation"",""input_modalities"":""text and image"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""Xiaomi MiMo official evaluation prompt"",""sampling"":""pass@1; source does not state trial count"",""temperature"":""1.0; top_p=0.95"",""tools"":""source-reported tools; exact tool set undisclosed"",""trials"":""source does not state""}",False,official_blog,verified,Locked Xiaomi observation for Humanity's Last Exam.
mimo-v2.5-pro,humaneval_plus,75.6,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""256K (reported score-table/model-download setting)"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""1-shot few-shot continuation"",""sampling"":""1-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The locked MiMo-V2.5-Pro-Base config exposes max_position_embeddings=1,048,576 while the score-table/download surface reports 256K; both exact values and the conflict are preserved."
mimo-v2.5-pro,livecodebench_v6,39.6,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""256K (reported score-table/model-download setting)"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""1-shot few-shot continuation"",""sampling"":""1-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The locked MiMo-V2.5-Pro-Base config exposes max_position_embeddings=1,048,576 while the score-table/download surface reports 256K; both exact values and the conflict are preserved."
mimo-v2.5-pro,math,86.2,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""256K (reported score-table/model-download setting)"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""4-shot few-shot continuation"",""sampling"":""4-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The locked MiMo-V2.5-Pro-Base config exposes max_position_embeddings=1,048,576 while the score-table/download surface reports 256K; both exact values and the conflict are preserved."
mimo-v2.5-pro,mbpp_plus,74.1,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""256K (reported score-table/model-download setting)"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""3-shot few-shot continuation"",""sampling"":""3-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The locked MiMo-V2.5-Pro-Base config exposes max_position_embeddings=1,048,576 while the score-table/download surface reports 256K; both exact values and the conflict are preserved."
mimo-v2.5-pro,mmlu,89.4,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""256K (reported score-table/model-download setting)"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""5-shot few-shot continuation"",""sampling"":""5-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The locked MiMo-V2.5-Pro-Base config exposes max_position_embeddings=1,048,576 while the score-table/download surface reports 256K; both exact values and the conflict are preserved."
mimo-v2.5-pro,mmlu_pro,68.5,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""256K (reported score-table/model-download setting)"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""5-shot few-shot continuation"",""sampling"":""5-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The locked MiMo-V2.5-Pro-Base config exposes max_position_embeddings=1,048,576 while the score-table/download surface reports 256K; both exact values and the conflict are preserved."
mimo-v2.5-pro,mmlu_redux,92.8,https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/resolve/21d1ecfecd7bd70f31be25ca49d7edd21f003659/README.md,"{""checkpoint"":""base"",""context"":""256K (reported score-table/model-download setting)"",""effort"":""not applicable"",""harness"":""Xiaomi MiMo base-checkpoint evaluation"",""input_modalities"":""text"",""judge"":""benchmark-native rule-based evaluator"",""mode"":""base pretrained checkpoint"",""prompt_style"":""5-shot few-shot continuation"",""sampling"":""5-shot; single completion per prompt"",""temperature"":""source does not state"",""tools"":""none"",""trials"":""one completion per item""}",False,official_model_card_base_checkpoint,verified,"Base checkpoint observation attached to the canonical post-trained model ID. Base result is attached to the canonical post-trained model ID. The locked MiMo-V2.5-Pro-Base config exposes max_position_embeddings=1,048,576 while the score-table/download surface reports 256K; both exact values and the conflict are preserved."
mimo-v2.5-pro,swe_bench_pro,57.2,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""1,048,576"",""effort"":""default"",""harness"":""official Xiaomi MiMo family benchmark evaluation"",""input_modalities"":""text"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""Xiaomi MiMo official evaluation prompt"",""sampling"":""source does not state repeated-trial count"",""temperature"":""1.0; top_p=0.95"",""tools"":""agentic code/terminal tools"",""trials"":""source does not state""}",False,official_blog,verified,Locked Xiaomi observation for SWE-Bench Pro.
mimo-v2.5-pro,swe_bench_verified,78.9,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""1,048,576"",""effort"":""default"",""harness"":""official Xiaomi MiMo family benchmark evaluation"",""input_modalities"":""text"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""Xiaomi MiMo official evaluation prompt"",""sampling"":""source does not state repeated-trial count"",""temperature"":""1.0; top_p=0.95"",""tools"":""agentic code/terminal tools"",""trials"":""source does not state""}",False,official_blog,verified,Locked Xiaomi observation for SWE-bench Verified.
mimo-v2.5-pro,tau3_bench,72.9,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""1,048,576"",""effort"":""default"",""harness"":""official Xiaomi MiMo family benchmark evaluation"",""input_modalities"":""text"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""Xiaomi MiMo official evaluation prompt"",""sampling"":""pass@1; source does not state repeated-trial count"",""temperature"":""1.0; top_p=0.95"",""tools"":""none"",""trials"":""source does not state""}",False,official_blog,verified,Locked Xiaomi observation for τ³-bench.
mimo-v2.5-pro,terminal_bench,68.4,https://mimo.xiaomi.com/mimo-v2-5-pro,"{""context"":""1,048,576"",""effort"":""default"",""harness"":""official Xiaomi MiMo family benchmark evaluation"",""input_modalities"":""text"",""judge"":""benchmark-specified"",""mode"":""thinking"",""prompt_style"":""Xiaomi MiMo official evaluation prompt"",""sampling"":""source does not state repeated-trial count"",""temperature"":""1.0; top_p=0.95"",""tools"":""agentic code/terminal tools"",""trials"":""source does not state""}",False,official_blog,verified,Locked Xiaomi observation for Terminal-Bench 2.0.
minimax-m2.7,claw_eval_pass3,49.7,https://raw.githubusercontent.com/claw-eval/claw-eval.github.io/19d378a646d892ca6c8148691aea02da1b184eb8/assets/benchmark-0fIVfCD7.js,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official Claw-Eval v1.1 leaderboard"",""input_modalities"":""text"",""judge"":""full-trajectory benchmark grading"",""mode"":""source does not state"",""prompt_style"":""official Claw-Eval task prompts"",""sampling"":""Pass^3; all three trajectories must pass"",""task_count"":161,""task_subset"":""core-general 161-task subset"",""temperature"":""source does not state"",""tools"":""official Claw-Eval agent environment"",""trials"":""3 per task; 483 generations""}",False,official_blog,verified,"Average tokens per trajectory=91135 is a physical efficiency annotation, not a benchmark score or score setting. Exact official leaderboard field for the core-general 161-task subset. Model mode/effort is not inferred."
muse-spark,claw_eval_pass3,62.7,https://raw.githubusercontent.com/claw-eval/claw-eval.github.io/19d378a646d892ca6c8148691aea02da1b184eb8/assets/benchmark-0fIVfCD7.js,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official Claw-Eval v1.1 leaderboard"",""input_modalities"":""text"",""judge"":""full-trajectory benchmark grading"",""mode"":""source does not state"",""prompt_style"":""official Claw-Eval task prompts"",""sampling"":""Pass^3; all three trajectories must pass"",""task_count"":161,""task_subset"":""core-general 161-task subset"",""temperature"":""source does not state"",""tools"":""official Claw-Eval agent environment"",""trials"":""3 per task; 483 generations""}",False,official_blog,verified,"Average tokens per trajectory=162688 is a physical efficiency annotation, not a benchmark score or score setting. Exact official leaderboard field for the core-general 161-task subset. Model mode/effort is not inferred."
qwen3.5-397b,claw_eval_pass3,57.8,https://raw.githubusercontent.com/claw-eval/claw-eval.github.io/19d378a646d892ca6c8148691aea02da1b184eb8/assets/benchmark-0fIVfCD7.js,"{""context"":""source does not state"",""effort"":""source does not state"",""harness"":""official Claw-Eval v1.1 leaderboard"",""input_modalities"":""text"",""judge"":""full-trajectory benchmark grading"",""mode"":""source does not state"",""prompt_style"":""official Claw-Eval task prompts"",""sampling"":""Pass^3; all three trajectories must pass"",""task_count"":161,""task_subset"":""core-general 161-task subset"",""temperature"":""source does not state"",""tools"":""official Claw-Eval agent environment"",""trials"":""3 per task; 483 generations""}",False,official_blog,verified,"Average tokens per trajectory=147963 is a physical efficiency annotation, not a benchmark score or score setting. Exact official leaderboard field for the core-general 161-task subset. Model mode/effort is not inferred."
ling-2.6-1t,aime_2026,87.4,https://arxiv.org/pdf/2606.15079v1,"{""aggregation"":""mean@64"",""checkpoint"":""post-trained"",""context_length"":""262,144 native config"",""context_management"":""none stated"",""dataset_version"":""AIME 2026"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""benchmark-specified"",""max_new_tokens"":""source does not state"",""mode"":""instant/fast-thinking"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":64,""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":64}",True,official_paper,verified,Displayed exactly as 87.40. Locked official InclusionAI observation.
ling-2.6-1t,bfcl_v4,70.64,https://arxiv.org/pdf/2606.15079v1,"{""aggregation"":""Accuracy"",""checkpoint"":""post-trained"",""context_length"":""262,144 native config"",""context_management"":""none stated"",""dataset_version"":""BFCL v4"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""Berkeley Function Calling Leaderboard"",""judge"":""official function-call match"",""max_new_tokens"":""source does not state"",""mode"":""instant/fast-thinking"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":1,""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""benchmark-provided functions"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":1}",True,official_paper,verified,Displayed exactly as 70.64. Locked official InclusionAI observation.
ling-2.6-1t,bigbench_extra_hard,52.37,https://arxiv.org/pdf/2606.15079v1,"{""aggregation"":""pass@1"",""checkpoint"":""post-trained"",""context_length"":""262,144 native config"",""context_management"":""none stated"",""dataset_version"":""bbeh"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""official BBEH exact-match/micro-accuracy evaluator"",""max_new_tokens"":""source does not state"",""mode"":""instant/fast-thinking"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":1,""search"":""none"",""shots"":""source does not state"",""split"":""full 4520 examples"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":1}",True,official_paper,verified,Displayed exactly as 52.37. Locked official InclusionAI observation.
ling-2.6-1t,chinese_simpleqa,76.53,https://arxiv.org/pdf/2606.15079v1,"{""aggregation"":""Correct"",""checkpoint"":""post-trained"",""context_length"":""262,144 native config"",""context_management"":""none stated"",""dataset_version"":""C-SimpleQA"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""benchmark-specified"",""max_new_tokens"":""source does not state"",""mode"":""instant/fast-thinking"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":1,""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":1}",True,official_paper,verified,Displayed exactly as 76.53. Locked official InclusionAI observation.
ling-2.6-1t,gpqa_diamond,76.17,https://arxiv.org/pdf/2606.15079v1,"{""aggregation"":""mean@4"",""checkpoint"":""post-trained"",""context_length"":""262,144 native config"",""context_management"":""none stated"",""dataset_version"":""GPQA-Diamond"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""benchmark-specified"",""max_new_tokens"":""source does not state"",""mode"":""instant/fast-thinking"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":4,""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":4}",True,official_paper,verified,Displayed exactly as 76.17. Locked official InclusionAI observation.
ling-2.6-1t,hmmt_nov_2025,81.93,https://arxiv.org/pdf/2606.15079v1,"{""aggregation"":""mean@64"",""checkpoint"":""post-trained"",""context_length"":""262,144 native config"",""context_management"":""none stated"",""dataset_version"":""HMMT November 2025"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""benchmark-specified"",""max_new_tokens"":""source does not state"",""mode"":""instant/fast-thinking"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":64,""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":64}",True,official_paper,verified,Displayed exactly as 81.93. Locked official InclusionAI observation.
ling-2.6-1t,ifbench,57.62,https://arxiv.org/pdf/2606.15079v1,"{""aggregation"":""source does not state"",""checkpoint"":""post-trained"",""context_length"":""262,144 native config"",""context_management"":""none stated"",""dataset_version"":""IFBench"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""benchmark-specified"",""max_new_tokens"":""source does not state"",""mode"":""instant/fast-thinking"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":""source does not state"",""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":""source does not state""}",True,official_paper,verified,Displayed exactly as 57.62. Locked official InclusionAI observation.
ling-2.6-1t,imo_answerbench,65.81,https://arxiv.org/pdf/2606.15079v1,"{""aggregation"":""mean@8"",""checkpoint"":""post-trained"",""context_length"":""262,144 native config"",""context_management"":""none stated"",""dataset_version"":""IMO-AnswerBench"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""benchmark-specified"",""max_new_tokens"":""source does not state"",""mode"":""instant/fast-thinking"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":8,""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":8}",True,official_paper,verified,Displayed exactly as 65.81. Locked official InclusionAI observation.
ling-2.6-1t,livecodebench_v6_2408_2505,65.58,https://arxiv.org/pdf/2606.15079v1,"{""aggregation"":""mean@4"",""checkpoint"":""post-trained"",""context_length"":""262,144 native config"",""context_management"":""none stated"",""dataset_version"":""LiveCodeBench v6, 2024-08 to 2025-05"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""official test execution"",""max_new_tokens"":""source does not state"",""mode"":""instant/fast-thinking"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":4,""search"":""none"",""shots"":""source does not state"",""split"":""454 code-generation problems"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""code execution for generated solutions"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":4}",True,official_paper,verified,Displayed exactly as 65.58. Locked official InclusionAI observation.
ling-2.6-1t,longbench_v2,48.31,https://arxiv.org/pdf/2606.15079v1,"{""aggregation"":""source does not state"",""checkpoint"":""post-trained"",""context_length"":""262,144 native config"",""context_management"":""none stated"",""dataset_version"":""LongBenchV2"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""benchmark-specified"",""max_new_tokens"":""source does not state"",""mode"":""instant/fast-thinking"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":""source does not state"",""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":""source does not state""}",True,official_paper,verified,Displayed exactly as 48.31. Locked official InclusionAI observation.
ling-2.6-1t,simpleqa_verified,31.5,https://arxiv.org/pdf/2606.15079v1,"{""aggregation"":""Correct"",""checkpoint"":""post-trained"",""context_length"":""262,144 native config"",""context_management"":""none stated"",""dataset_version"":""SimpleQA-Verified"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""benchmark-specified"",""max_new_tokens"":""source does not state"",""mode"":""instant/fast-thinking"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":1,""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":1}",True,official_paper,verified,Displayed exactly as 31.50. Locked official InclusionAI observation.
ling-2.6-1t,supergpqa,58.32,https://arxiv.org/pdf/2606.15079v1,"{""aggregation"":""EM, CoT"",""checkpoint"":""post-trained"",""context_length"":""262,144 native config"",""context_management"":""none stated"",""dataset_version"":""SuperGPQA"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""benchmark-specified"",""max_new_tokens"":""source does not state"",""mode"":""instant/fast-thinking"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":1,""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":1}",True,official_paper,verified,Displayed exactly as 58.32. Locked official InclusionAI observation.
ling-2.6-1t,swe_bench_verified,72.2,https://huggingface.co/inclusionAI/Ling-2.6-1T/blob/07b9e26bb1f97494b4fd62eed83f22722ffd560a/.eval_results/swe-bench_verified.yaml,"{""aggregation"":""% resolved (pass@1)"",""checkpoint"":""post-trained"",""context_length"":""262,144 native config"",""context_management"":""none stated"",""dataset_version"":""SWE-bench Verified"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""source does not state"",""judge"":""official repository test suite"",""max_new_tokens"":""source does not state"",""mode"":""instant/fast-thinking"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":1,""search"":""none"",""shots"":""source does not state"",""split"":""Verified 500"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""coding-agent shell/editor tools"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":1}",True,official_model_card,verified,Displayed exactly as 72.2. Locked official InclusionAI observation.
ling-2.6-1t,tau2_bench_avg,78.36,https://arxiv.org/pdf/2606.15079v1,"{""aggregation"":""source-reported average"",""checkpoint"":""post-trained"",""context_length"":""262,144 native config"",""context_management"":""none stated"",""dataset_version"":""tau2 v1.0.0 average"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official tau2 harness with source-specific user simulator"",""judge"":""task success"",""max_new_tokens"":""source does not state"",""mode"":""instant/fast-thinking"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":""source does not state"",""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""official tau2 domain tools"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":""source does not state""}",True,official_paper,verified,Displayed exactly as 78.36. Locked official InclusionAI observation.
ling-2.6-1t,terminal_bench,40.45,https://arxiv.org/pdf/2606.15079v1,"{""aggregation"":""Accuracy"",""checkpoint"":""post-trained"",""context_length"":""262,144 native config"",""context_management"":""none stated"",""dataset_version"":""Terminal-Bench 2.0, 74 tasks"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""source-specified terminal agent"",""judge"":""programmatic task tests"",""max_new_tokens"":""source does not state"",""mode"":""instant/fast-thinking"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":1,""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""terminal"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":1}",True,official_paper,verified,Displayed exactly as 40.45. Locked official InclusionAI observation.
ling-2.6-flash,aime_2026,73.85,https://huggingface.co/inclusionAI/Ling-2.6-flash/blob/11236968749136a78d4b7cbfb786c4460f2f353e/.eval_results/aime_2026.yaml,"{""aggregation"":""mean@64"",""checkpoint"":""post-trained"",""context_length"":""131,072 native config"",""context_management"":""none stated"",""dataset_version"":""AIME 2026"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""benchmark-specified"",""max_new_tokens"":""source does not state"",""mode"":""instant/non-reasoning"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":64,""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":64}",True,official_model_card,verified,Displayed exactly as 73.85. Locked official InclusionAI observation.
ling-2.6-flash,bfcl_v3,76.05,https://raw.githubusercontent.com/inclusionAI/LLaDA2.X/f402f0b52817e1c3586a024ebf89e69ad8ca5523/LLaDA2_2_tech_report.pdf,"{""aggregation"":""source does not state"",""checkpoint"":""post-trained"",""context_length"":""131,072 native config"",""context_management"":""none stated"",""dataset_version"":""BFCL v3"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""Berkeley Function Calling Leaderboard"",""judge"":""official function-call match"",""max_new_tokens"":""source does not state"",""mode"":""instant/non-reasoning"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":""source does not state"",""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""benchmark-provided functions"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":""source does not state""}",True,official_paper,verified,Displayed exactly as 76.05. Locked official InclusionAI observation.
ling-2.6-flash,bfcl_v4,66.81,https://arxiv.org/pdf/2606.15079v1,"{""aggregation"":""Accuracy"",""checkpoint"":""post-trained"",""context_length"":""131,072 native config"",""context_management"":""none stated"",""dataset_version"":""BFCL v4"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""Berkeley Function Calling Leaderboard"",""judge"":""official function-call match"",""max_new_tokens"":""source does not state"",""mode"":""instant/non-reasoning"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":1,""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""benchmark-provided functions"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":1}",True,official_paper,verified,Displayed exactly as 66.81. Locked official InclusionAI observation.
ling-2.6-flash,chinese_simpleqa,60.23,https://arxiv.org/pdf/2606.15079v1,"{""aggregation"":""Correct"",""checkpoint"":""post-trained"",""context_length"":""131,072 native config"",""context_management"":""none stated"",""dataset_version"":""C-SimpleQA"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""benchmark-specified"",""max_new_tokens"":""source does not state"",""mode"":""instant/non-reasoning"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":1,""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":1}",True,official_paper,verified,Displayed exactly as 60.23. Locked official InclusionAI observation.
ling-2.6-flash,gpqa_diamond,60.35,https://raw.githubusercontent.com/inclusionAI/LLaDA2.X/f402f0b52817e1c3586a024ebf89e69ad8ca5523/LLaDA2_2_tech_report.pdf,"{""aggregation"":""source does not state"",""checkpoint"":""post-trained"",""context_length"":""131,072 native config"",""context_management"":""none stated"",""dataset_version"":""GPQA-Diamond"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""benchmark-specified"",""max_new_tokens"":""source does not state"",""mode"":""instant/non-reasoning"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":""source does not state"",""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":""source does not state""}",True,official_paper,verified,Displayed exactly as 60.35. Locked official InclusionAI observation.
ling-2.6-flash,hmmt_feb_2026,49.29,https://huggingface.co/inclusionAI/Ling-2.6-flash/blob/11236968749136a78d4b7cbfb786c4460f2f353e/.eval_results/hmmt_feb_2026.yaml,"{""aggregation"":""mean@64"",""checkpoint"":""post-trained"",""context_length"":""131,072 native config"",""context_management"":""none stated"",""dataset_version"":""HMMT February 2026"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""benchmark-specified"",""max_new_tokens"":""source does not state"",""mode"":""instant/non-reasoning"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":64,""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":64}",True,official_model_card,verified,Displayed exactly as 49.29. Locked official InclusionAI observation.
ling-2.6-flash,ifbench,57.4,https://arxiv.org/pdf/2606.15079v1,"{""aggregation"":""mean@5"",""checkpoint"":""post-trained"",""context_length"":""131,072 native config"",""context_management"":""none stated"",""dataset_version"":""IFBench"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""benchmark-specified"",""max_new_tokens"":""source does not state"",""mode"":""instant/non-reasoning"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":5,""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":5}",True,official_paper,verified,Displayed exactly as 57.40. Locked official InclusionAI observation.
ling-2.6-flash,imo_answerbench,54.28,https://arxiv.org/pdf/2606.15079v1,"{""aggregation"":""mean@8"",""checkpoint"":""post-trained"",""context_length"":""131,072 native config"",""context_management"":""none stated"",""dataset_version"":""IMO-AnswerBench"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""benchmark-specified"",""max_new_tokens"":""source does not state"",""mode"":""instant/non-reasoning"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":8,""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":8}",True,official_paper,verified,Displayed exactly as 54.28. Locked official InclusionAI observation.
ling-2.6-flash,korbench,66.48,https://raw.githubusercontent.com/inclusionAI/LLaDA2.X/f402f0b52817e1c3586a024ebf89e69ad8ca5523/LLaDA2_2_tech_report.pdf,"{""aggregation"":""source does not state"",""checkpoint"":""post-trained"",""context_length"":""131,072 native config"",""context_management"":""none stated"",""dataset_version"":""KORBench"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""benchmark-specified"",""max_new_tokens"":""source does not state"",""mode"":""instant/non-reasoning"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":""source does not state"",""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":""source does not state""}",True,official_paper,verified,Displayed exactly as 66.48. Locked official InclusionAI observation.
ling-2.6-flash,livecodebench_v6_2408_2505,62.28,https://arxiv.org/pdf/2606.15079v1,"{""aggregation"":""mean@4"",""checkpoint"":""post-trained"",""context_length"":""131,072 native config"",""context_management"":""none stated"",""dataset_version"":""LiveCodeBench v6, 2024-08 to 2025-05"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""official test execution"",""max_new_tokens"":""source does not state"",""mode"":""instant/non-reasoning"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":4,""search"":""none"",""shots"":""source does not state"",""split"":""454 code-generation problems"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""code execution for generated solutions"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":4}",True,official_paper,verified,Displayed exactly as 62.28. Locked official InclusionAI observation.
ling-2.6-flash,longbench_v2,42.94,https://raw.githubusercontent.com/inclusionAI/LLaDA2.X/f402f0b52817e1c3586a024ebf89e69ad8ca5523/LLaDA2_2_tech_report.pdf,"{""aggregation"":""source does not state"",""checkpoint"":""post-trained"",""context_length"":""131,072 native config"",""context_management"":""none stated"",""dataset_version"":""LongBench v2"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""benchmark-specified"",""max_new_tokens"":""source does not state"",""mode"":""instant/non-reasoning"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":""source does not state"",""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":""source does not state""}",True,official_paper,verified,Displayed exactly as 42.94. Locked official InclusionAI observation.
ling-2.6-flash,mcpatlas,41.12,https://raw.githubusercontent.com/inclusionAI/LLaDA2.X/f402f0b52817e1c3586a024ebf89e69ad8ca5523/LLaDA2_2_tech_report.pdf,"{""aggregation"":""arithmetic mean of 5 runs"",""checkpoint"":""post-trained"",""context_length"":""131,072 native config"",""context_management"":""none stated"",""dataset_version"":""MCP-Atlas"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official MCP-Atlas v1, 20-turn limit"",""judge"":""Gemini-2.5-Pro claim coverage; pass at >=0.75"",""max_new_tokens"":""source does not state"",""mode"":""instant/non-reasoning"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":1,""search"":""none"",""shots"":""source does not state"",""split"":""public 500 tasks"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""36 MCP servers / public tool suite"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":5}",True,official_paper,verified,Displayed exactly as 41.12. Locked official InclusionAI observation.
ling-2.6-flash,multi_if,74.8,https://arxiv.org/pdf/2606.15079v1,"{""aggregation"":""turn-3"",""checkpoint"":""post-trained"",""context_length"":""131,072 native config"",""context_management"":""none stated"",""dataset_version"":""Multi-IF turn-3"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""benchmark-specified"",""max_new_tokens"":""source does not state"",""mode"":""instant/non-reasoning"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":1,""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":1}",True,official_paper,verified,Displayed exactly as 74.80. Locked official InclusionAI observation.
ling-2.6-flash,multichallenge,39.71,https://arxiv.org/pdf/2606.15079v1,"{""aggregation"":""Accuracy"",""checkpoint"":""post-trained"",""context_length"":""131,072 native config"",""context_management"":""none stated"",""dataset_version"":""Multichallenge"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""benchmark-specified"",""max_new_tokens"":""source does not state"",""mode"":""instant/non-reasoning"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":1,""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":1}",True,official_paper,verified,Displayed exactly as 39.71. Locked official InclusionAI observation.
ling-2.6-flash,simpleqa_verified,15.1,https://arxiv.org/pdf/2606.15079v1,"{""aggregation"":""Correct"",""checkpoint"":""post-trained"",""context_length"":""131,072 native config"",""context_management"":""none stated"",""dataset_version"":""SimpleQA-Verified"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""benchmark-specified"",""max_new_tokens"":""source does not state"",""mode"":""instant/non-reasoning"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":1,""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":1}",True,official_paper,verified,Displayed exactly as 15.10. Locked official InclusionAI observation.
ling-2.6-flash,swe_bench_multilingual,33.73,https://raw.githubusercontent.com/inclusionAI/LLaDA2.X/f402f0b52817e1c3586a024ebf89e69ad8ca5523/LLaDA2_2_tech_report.pdf,"{""aggregation"":""source does not state"",""checkpoint"":""post-trained"",""context_length"":""131,072 native config"",""context_management"":""none stated"",""dataset_version"":""SWE-bench Multilingual"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""Claude Code"",""judge"":""official repository test suite"",""max_new_tokens"":""source does not state"",""mode"":""instant/non-reasoning"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":""source does not state"",""search"":""none"",""shots"":""source does not state"",""split"":""Multilingual 300"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""coding-agent shell/editor tools"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":""source does not state""}",True,official_paper,verified,Displayed exactly as 33.73. Locked official InclusionAI observation.
ling-2.6-flash,swe_bench_pro,31.88,https://raw.githubusercontent.com/inclusionAI/LLaDA2.X/f402f0b52817e1c3586a024ebf89e69ad8ca5523/LLaDA2_2_tech_report.pdf,"{""aggregation"":""source does not state"",""checkpoint"":""post-trained"",""context_length"":""131,072 native config"",""context_management"":""none stated"",""dataset_version"":""SWE-bench Pro"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""Claude Code"",""judge"":""official repository test suite"",""max_new_tokens"":""source does not state"",""mode"":""instant/non-reasoning"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":""source does not state"",""search"":""none"",""shots"":""source does not state"",""split"":""public 731"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""coding-agent shell/editor tools"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":""source does not state""}",True,official_paper,verified,Displayed exactly as 31.88. Locked official InclusionAI observation.
ling-2.6-flash,swe_bench_verified,61.2,https://huggingface.co/inclusionAI/Ling-2.6-flash/blob/11236968749136a78d4b7cbfb786c4460f2f353e/.eval_results/swe-bench_verified.yaml,"{""aggregation"":""% resolved (pass@1)"",""checkpoint"":""post-trained"",""context_length"":""131,072 native config"",""context_management"":""none stated"",""dataset_version"":""SWE-bench Verified"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""OpenHands"",""judge"":""official repository test suite"",""max_new_tokens"":""source does not state"",""mode"":""instant/non-reasoning"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":1,""search"":""none"",""shots"":""source does not state"",""split"":""Verified 500"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""coding-agent shell/editor tools"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":1}",True,official_model_card,verified,Displayed exactly as 61.2. Locked official InclusionAI observation.
ling-2.6-flash,tau2_bench_airline,63.0,https://mdn.alipayobjects.com/huamei_3p6pd0/afts/img/4bI1SK8pNM8AAAAAgBAAAAgADryCAQFr/original,"{""aggregation"":""source does not state"",""checkpoint"":""post-trained"",""context_length"":""131,072 native config"",""context_management"":""none stated"",""dataset_version"":""tau2 v1.0.0 airline"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official tau2 harness with source-specific user simulator"",""judge"":""task success"",""max_new_tokens"":""source does not state"",""mode"":""instant/non-reasoning"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":""source does not state"",""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""official tau2 domain tools"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":""source does not state""}",True,official_model_card,verified,Displayed exactly as 63.00. Locked official InclusionAI observation.
ling-2.6-flash,tau2_bench_avg,76.36,https://arxiv.org/pdf/2606.15079v1,"{""aggregation"":""mean@4"",""checkpoint"":""post-trained"",""context_length"":""131,072 native config"",""context_management"":""none stated"",""dataset_version"":""tau2 v1.0.0 average"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official tau2 harness with source-specific user simulator"",""judge"":""GPT-5.2 user simulator; task-success judge"",""max_new_tokens"":""source does not state"",""mode"":""instant/non-reasoning"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":4,""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""official tau2 domain tools"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":4}",True,official_paper,verified,Displayed exactly as 76.36. Locked official InclusionAI observation.
ling-2.6-flash,tau2_bench_retail,71.49,https://mdn.alipayobjects.com/huamei_3p6pd0/afts/img/4bI1SK8pNM8AAAAAgBAAAAgADryCAQFr/original,"{""aggregation"":""source does not state"",""checkpoint"":""post-trained"",""context_length"":""131,072 native config"",""context_management"":""none stated"",""dataset_version"":""tau2 v1.0.0 retail"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official tau2 harness with source-specific user simulator"",""judge"":""task success"",""max_new_tokens"":""source does not state"",""mode"":""instant/non-reasoning"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":""source does not state"",""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""official tau2 domain tools"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":""source does not state""}",True,official_model_card,verified,Displayed exactly as 71.49. Locked official InclusionAI observation.
ling-2.6-flash,tau2_bench_telecom,94.96,https://mdn.alipayobjects.com/huamei_3p6pd0/afts/img/4bI1SK8pNM8AAAAAgBAAAAgADryCAQFr/original,"{""aggregation"":""source does not state"",""checkpoint"":""post-trained"",""context_length"":""131,072 native config"",""context_management"":""none stated"",""dataset_version"":""tau2 v1.0.0 telecom"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official tau2 harness with source-specific user simulator"",""judge"":""task success"",""max_new_tokens"":""source does not state"",""mode"":""instant/non-reasoning"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":""source does not state"",""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""official tau2 domain tools"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":""source does not state""}",True,official_model_card,verified,Displayed exactly as 94.96. Locked official InclusionAI observation.
ling-3.0-flash,aa_lcr,65.1,https://intranetproxy.alipay.com/skylark/lark/0/2026/png/23157180/1785831264180-d6ca4404-acef-4424-84db-fbc5a4c6db5f.png,"{""aggregation"":""source does not state"",""checkpoint"":""post-trained"",""context_length"":""262,144 native evaluation/config"",""context_management"":""none stated"",""dataset_version"":""AA-LCR Intelligence Index methodology"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""Artificial Analysis equality-checker LLM"",""max_new_tokens"":""source does not state"",""mode"":""thinking enabled by default"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":""source does not state"",""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""0.6"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""20"",""top_p"":""0.95"",""trials"":""source does not state""}",True,official_model_card,verified,Displayed exactly as 65.1. Locked official InclusionAI observation.
ling-3.0-flash,aime_2026,93.2,https://huggingface.co/inclusionAI/Ling-3.0-flash/blob/ecde16176a497adaff7419ff4de59da603c4edaa/.eval_results/aime_2026.yaml,"{""aggregation"":""source does not state"",""checkpoint"":""post-trained"",""context_length"":""262,144 native evaluation/config"",""context_management"":""none stated"",""dataset_version"":""AIME 2026"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""benchmark-specified"",""max_new_tokens"":""source does not state"",""mode"":""thinking enabled by default"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":""source does not state"",""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""0.6"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""20"",""top_p"":""0.95"",""trials"":""source does not state""}",True,official_model_card,verified,Displayed exactly as 93.2. Locked official InclusionAI observation.
ling-3.0-flash,bfcl_v4,73.0,https://intranetproxy.alipay.com/skylark/lark/0/2026/png/23157180/1785831264180-d6ca4404-acef-4424-84db-fbc5a4c6db5f.png,"{""aggregation"":""source does not state"",""checkpoint"":""post-trained"",""context_length"":""262,144 native evaluation/config"",""context_management"":""none stated"",""dataset_version"":""BFCL v4"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""Berkeley Function Calling Leaderboard"",""judge"":""official function-call match"",""max_new_tokens"":""source does not state"",""mode"":""thinking enabled by default"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":""source does not state"",""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""0.6"",""timeout"":""source does not state"",""tools"":""benchmark-provided functions"",""top_k"":""20"",""top_p"":""0.95"",""trials"":""source does not state""}",True,official_model_card,verified,Displayed exactly as 73.0. Locked official InclusionAI observation.
ling-3.0-flash,browsecomp,72.2,https://intranetproxy.alipay.com/skylark/lark/0/2026/png/23157180/1785831264180-d6ca4404-acef-4424-84db-fbc5a4c6db5f.png,"{""aggregation"":""source does not state"",""checkpoint"":""post-trained"",""context_length"":""262,144 native evaluation/config"",""context_management"":""single-agent resume from trajectory summary at 64K"",""dataset_version"":""BrowseComp"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""InclusionAI internal search-agent harness"",""judge"":""BrowseComp exact-answer judge"",""max_new_tokens"":""source does not state"",""mode"":""thinking enabled by default"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":""source does not state"",""search"":""enabled"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""0.6"",""timeout"":""source does not state"",""tools"":""browser/search and code execution"",""top_k"":""20"",""top_p"":""0.95"",""trials"":""source does not state""}",True,official_model_card,verified,Displayed exactly as 72.2 (w/ ctx); 82.0 (MA). Locked official InclusionAI observation.
ling-3.0-flash,draco,70.4,https://intranetproxy.alipay.com/skylark/lark/0/2026/png/23157180/1785831264180-d6ca4404-acef-4424-84db-fbc5a4c6db5f.png,"{""aggregation"":""source does not state"",""checkpoint"":""post-trained"",""context_length"":""262,144 native evaluation/config"",""context_management"":""none stated"",""dataset_version"":""Draco"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""InclusionAI internal search-agent harness"",""judge"":""official rubrics; Claude Opus 4.6 scoring model"",""max_new_tokens"":""source does not state"",""mode"":""thinking enabled by default"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":""source does not state"",""search"":""enabled"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""0.6"",""timeout"":""source does not state"",""tools"":""browser/search and code execution"",""top_k"":""20"",""top_p"":""0.95"",""trials"":""source does not state""}",True,official_model_card,verified,Displayed exactly as 70.4. Locked official InclusionAI observation.
ling-3.0-flash,gdpval_aa_v2_elo,1107.0,https://intranetproxy.alipay.com/skylark/lark/0/2026/png/23157180/1785831264180-d6ca4404-acef-4424-84db-fbc5a4c6db5f.png,"{""aggregation"":""source does not state"",""checkpoint"":""post-trained"",""context_length"":""262,144 native evaluation/config"",""context_management"":""none stated"",""dataset_version"":""GDPval v2-AA"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""GDPval-AA v2; 250 turns; 5-hour timeout"",""judge"":""pairwise judge panel / Elo"",""max_new_tokens"":""source does not state"",""mode"":""thinking enabled by default"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":""source does not state"",""search"":""none"",""shots"":""source does not state"",""split"":""220 public tasks"",""temperature"":""0.6"",""timeout"":""source does not state"",""tools"":""Stirrup agent tools"",""top_k"":""20"",""top_p"":""0.95"",""trials"":""source does not state""}",True,official_model_card,verified,Displayed exactly as 1107. Locked official InclusionAI observation.
ling-3.0-flash,hmmt_feb_2026,87.0,https://huggingface.co/inclusionAI/Ling-3.0-flash/blob/ecde16176a497adaff7419ff4de59da603c4edaa/.eval_results/hmmt_feb_2026.yaml,"{""aggregation"":""source does not state"",""checkpoint"":""post-trained"",""context_length"":""262,144 native evaluation/config"",""context_management"":""none stated"",""dataset_version"":""HMMT February 2026"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""benchmark-specified"",""max_new_tokens"":""source does not state"",""mode"":""thinking enabled by default"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":""source does not state"",""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""0.6"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""20"",""top_p"":""0.95"",""trials"":""source does not state""}",True,official_model_card,verified,Displayed exactly as 87.0. Locked official InclusionAI observation.
ling-3.0-flash,ifbench,74.5,https://intranetproxy.alipay.com/skylark/lark/0/2026/png/23157180/1785831264180-d6ca4404-acef-4424-84db-fbc5a4c6db5f.png,"{""aggregation"":""source does not state"",""checkpoint"":""post-trained"",""context_length"":""262,144 native evaluation/config"",""context_management"":""none stated"",""dataset_version"":""IFBench"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""benchmark-specified"",""max_new_tokens"":""source does not state"",""mode"":""thinking enabled by default"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":""source does not state"",""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""0.6"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""20"",""top_p"":""0.95"",""trials"":""source does not state""}",True,official_model_card,verified,Displayed exactly as 74.5. Locked official InclusionAI observation.
ling-3.0-flash,imo_answerbench,83.7,https://intranetproxy.alipay.com/skylark/lark/0/2026/png/23157180/1785831264180-d6ca4404-acef-4424-84db-fbc5a4c6db5f.png,"{""aggregation"":""source does not state"",""checkpoint"":""post-trained"",""context_length"":""262,144 native evaluation/config"",""context_management"":""none stated"",""dataset_version"":""IMO-AnswerBench"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""benchmark-specified"",""max_new_tokens"":""source does not state"",""mode"":""thinking enabled by default"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":""source does not state"",""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""0.6"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""20"",""top_p"":""0.95"",""trials"":""source does not state""}",True,official_model_card,verified,Displayed exactly as 83.7. Locked official InclusionAI observation.
ling-3.0-flash,livecodebench_v6_2408_2505,82.8,https://intranetproxy.alipay.com/skylark/lark/0/2026/png/23157180/1785831264180-d6ca4404-acef-4424-84db-fbc5a4c6db5f.png,"{""aggregation"":""source does not state"",""checkpoint"":""post-trained"",""context_length"":""262,144 native evaluation/config"",""context_management"":""none stated"",""dataset_version"":""LiveCodeBench v6, 2024-08 to 2025-05"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""official test execution"",""max_new_tokens"":""source does not state"",""mode"":""thinking enabled by default"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":""source does not state"",""search"":""none"",""shots"":""source does not state"",""split"":""454 code-generation problems"",""temperature"":""0.6"",""timeout"":""source does not state"",""tools"":""code execution for generated solutions"",""top_k"":""20"",""top_p"":""0.95"",""trials"":""source does not state""}",True,official_model_card,verified,Displayed exactly as 82.8. Locked official InclusionAI observation.
ling-3.0-flash,mcpatlas,65.5,https://intranetproxy.alipay.com/skylark/lark/0/2026/png/23157180/1785831264180-d6ca4404-acef-4424-84db-fbc5a4c6db5f.png,"{""aggregation"":""source does not state"",""checkpoint"":""post-trained"",""context_length"":""262,144 native evaluation/config"",""context_management"":""none stated"",""dataset_version"":""MCP-Atlas"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official MCP-Atlas v1, 20-turn limit"",""judge"":""Gemini-2.5-Pro claim coverage; pass at >=0.75"",""max_new_tokens"":""source does not state"",""mode"":""thinking enabled by default"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":""source does not state"",""search"":""none"",""shots"":""source does not state"",""split"":""public 500 tasks"",""temperature"":""0.6"",""timeout"":""source does not state"",""tools"":""36 MCP servers / public tool suite"",""top_k"":""20"",""top_p"":""0.95"",""trials"":""source does not state""}",True,official_model_card,verified,Displayed exactly as 65.5. Locked official InclusionAI observation.
ling-3.0-flash,multi_if,87.7,https://intranetproxy.alipay.com/skylark/lark/0/2026/png/23157180/1785831264180-d6ca4404-acef-4424-84db-fbc5a4c6db5f.png,"{""aggregation"":""source does not state"",""checkpoint"":""post-trained"",""context_length"":""262,144 native evaluation/config"",""context_management"":""none stated"",""dataset_version"":""Multi-IF turn-3"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""benchmark-specified"",""max_new_tokens"":""source does not state"",""mode"":""thinking enabled by default"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":""source does not state"",""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""0.6"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""20"",""top_p"":""0.95"",""trials"":""source does not state""}",True,official_model_card,verified,Displayed exactly as 87.7. Locked official InclusionAI observation.
ling-3.0-flash,swe_bench_multilingual,72.4,https://huggingface.co/inclusionAI/Ling-3.0-flash/blob/ecde16176a497adaff7419ff4de59da603c4edaa/.eval_results/swe-bench_multilingual.yaml,"{""aggregation"":""source does not state"",""checkpoint"":""post-trained"",""context_length"":""256K"",""context_management"":""none stated"",""dataset_version"":""SWE-bench Multilingual"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""OpenHands with tailored prompts"",""judge"":""official repository test suite"",""max_new_tokens"":""32K"",""mode"":""thinking enabled by default"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""OpenHands tailored prompts"",""samples"":""source does not state"",""search"":""none"",""shots"":""source does not state"",""split"":""Multilingual 300"",""temperature"":""0.6"",""timeout"":""source does not state"",""tools"":""coding-agent shell/editor tools"",""top_k"":""20"",""top_p"":""0.95"",""trials"":""source does not state""}",True,official_model_card,verified,Displayed exactly as 72.4. Locked official InclusionAI observation.
ling-3.0-flash,swe_bench_pro,56.6,https://huggingface.co/inclusionAI/Ling-3.0-flash/blob/ecde16176a497adaff7419ff4de59da603c4edaa/.eval_results/swe-bench_pro.yaml,"{""aggregation"":""source does not state"",""checkpoint"":""post-trained"",""context_length"":""256K"",""context_management"":""none stated"",""dataset_version"":""SWE-bench Pro"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""OpenHands with tailored prompts"",""judge"":""official repository test suite"",""max_new_tokens"":""32K"",""mode"":""thinking enabled by default"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""OpenHands tailored prompts"",""samples"":""source does not state"",""search"":""none"",""shots"":""source does not state"",""split"":""public 731"",""temperature"":""0.6"",""timeout"":""source does not state"",""tools"":""coding-agent shell/editor tools"",""top_k"":""20"",""top_p"":""0.95"",""trials"":""source does not state""}",True,official_model_card,verified,Displayed exactly as 56.6. Locked official InclusionAI observation.
ling-3.0-flash,terminal_bench_2_1,57.0,https://intranetproxy.alipay.com/skylark/lark/0/2026/png/23157180/1785831264180-d6ca4404-acef-4424-84db-fbc5a4c6db5f.png,"{""aggregation"":""arithmetic mean of 3 runs"",""checkpoint"":""post-trained"",""context_length"":""256K"",""context_management"":""none stated"",""dataset_version"":""Terminal-Bench 2.1, 89 tasks"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""Terminus 2 under Artificial Analysis protocol"",""judge"":""programmatic task tests"",""max_new_tokens"":""32K"",""mode"":""thinking enabled by default"",""multimodal_inputs"":false,""output_parser"":""provided JSON parser in preserve-thinking mode"",""prompt_style"":""source does not state"",""samples"":1,""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""0.6"",""timeout"":""2 hours per task"",""tools"":""terminal"",""top_k"":""20"",""top_p"":""1.0"",""trials"":3}",True,official_model_card,verified,Displayed exactly as 57.0. Locked official InclusionAI observation. The adjacent benchmark-labeled methodology bullet is scoped row-wide across every displayed model column.
llada2.2-flash,aime_2026,62.24,https://raw.githubusercontent.com/inclusionAI/LLaDA2.X/f402f0b52817e1c3586a024ebf89e69ad8ca5523/LLaDA2_2_tech_report.pdf,"{""aggregation"":""arithmetic mean of five runs"",""checkpoint"":""post-trained"",""context_length"":""131,072 / 128K"",""context_management"":""none stated"",""dataset_version"":""AIME 2026"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""benchmark-specified"",""max_new_tokens"":""source does not state"",""mode"":""block-diffusion with Levenshtein editing"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""block_length=32; threshold=0.5; editing_threshold=0.0"",""samples"":5,""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""1.0"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":5}",True,official_paper,verified,Displayed exactly as 62.24. Locked official InclusionAI observation.
llada2.2-flash,bfcl_v3,67.17,https://raw.githubusercontent.com/inclusionAI/LLaDA2.X/f402f0b52817e1c3586a024ebf89e69ad8ca5523/LLaDA2_2_tech_report.pdf,"{""aggregation"":""arithmetic mean of five runs"",""checkpoint"":""post-trained"",""context_length"":""131,072 / 128K"",""context_management"":""none stated"",""dataset_version"":""BFCL v3"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""Berkeley Function Calling Leaderboard"",""judge"":""official function-call match"",""max_new_tokens"":""source does not state"",""mode"":""block-diffusion with Levenshtein editing"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""block_length=32; threshold=0.5; editing_threshold=0.0"",""samples"":5,""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""1.0"",""timeout"":""source does not state"",""tools"":""benchmark-provided functions"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":5}",True,official_paper,verified,Displayed exactly as 67.17. Locked official InclusionAI observation.
llada2.2-flash,bfcl_v4,60.78,https://raw.githubusercontent.com/inclusionAI/LLaDA2.X/f402f0b52817e1c3586a024ebf89e69ad8ca5523/LLaDA2_2_tech_report.pdf,"{""aggregation"":""arithmetic mean of five runs"",""checkpoint"":""post-trained"",""context_length"":""131,072 / 128K"",""context_management"":""none stated"",""dataset_version"":""BFCL v4"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""Berkeley Function Calling Leaderboard"",""judge"":""official function-call match"",""max_new_tokens"":""source does not state"",""mode"":""block-diffusion with Levenshtein editing"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""block_length=32; threshold=0.5; editing_threshold=0.0"",""samples"":5,""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""1.0"",""timeout"":""source does not state"",""tools"":""benchmark-provided functions"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":5}",True,official_paper,verified,Displayed exactly as 60.78. Locked official InclusionAI observation.
llada2.2-flash,gpqa_diamond,48.67,https://raw.githubusercontent.com/inclusionAI/LLaDA2.X/f402f0b52817e1c3586a024ebf89e69ad8ca5523/LLaDA2_2_tech_report.pdf,"{""aggregation"":""arithmetic mean of five runs"",""checkpoint"":""post-trained"",""context_length"":""131,072 / 128K"",""context_management"":""none stated"",""dataset_version"":""GPQA-Diamond"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""benchmark-specified"",""max_new_tokens"":""source does not state"",""mode"":""block-diffusion with Levenshtein editing"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""block_length=32; threshold=0.5; editing_threshold=0.0"",""samples"":5,""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""1.0"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":5}",True,official_paper,verified,Displayed exactly as 48.67. Locked official InclusionAI observation.
llada2.2-flash,ifbench,30.2,https://raw.githubusercontent.com/inclusionAI/LLaDA2.X/f402f0b52817e1c3586a024ebf89e69ad8ca5523/LLaDA2_2_tech_report.pdf,"{""aggregation"":""arithmetic mean of five runs"",""checkpoint"":""post-trained"",""context_length"":""131,072 / 128K"",""context_management"":""none stated"",""dataset_version"":""IFBench"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""benchmark-specified"",""max_new_tokens"":""source does not state"",""mode"":""block-diffusion with Levenshtein editing"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""block_length=32; threshold=0.5; editing_threshold=0.0"",""samples"":5,""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""1.0"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":5}",True,official_paper,verified,Displayed exactly as 30.20. Locked official InclusionAI observation.
llada2.2-flash,korbench,60.96,https://raw.githubusercontent.com/inclusionAI/LLaDA2.X/f402f0b52817e1c3586a024ebf89e69ad8ca5523/LLaDA2_2_tech_report.pdf,"{""aggregation"":""arithmetic mean of five runs"",""checkpoint"":""post-trained"",""context_length"":""131,072 / 128K"",""context_management"":""none stated"",""dataset_version"":""KORBench"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""benchmark-specified"",""max_new_tokens"":""source does not state"",""mode"":""block-diffusion with Levenshtein editing"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""block_length=32; threshold=0.5; editing_threshold=0.0"",""samples"":5,""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""1.0"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":5}",True,official_paper,verified,Displayed exactly as 60.96. Locked official InclusionAI observation.
llada2.2-flash,longbench_v2,45.13,https://raw.githubusercontent.com/inclusionAI/LLaDA2.X/f402f0b52817e1c3586a024ebf89e69ad8ca5523/LLaDA2_2_tech_report.pdf,"{""aggregation"":""arithmetic mean of five runs"",""checkpoint"":""post-trained"",""context_length"":""131,072 / 128K"",""context_management"":""none stated"",""dataset_version"":""LongBench v2"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""benchmark-specified"",""max_new_tokens"":""source does not state"",""mode"":""block-diffusion with Levenshtein editing"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""block_length=32; threshold=0.5; editing_threshold=0.0"",""samples"":5,""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""1.0"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":5}",True,official_paper,verified,Displayed exactly as 45.13. Locked official InclusionAI observation.
llada2.2-flash,mcpatlas,46.21,https://raw.githubusercontent.com/inclusionAI/LLaDA2.X/f402f0b52817e1c3586a024ebf89e69ad8ca5523/LLaDA2_2_tech_report.pdf,"{""aggregation"":""arithmetic mean of five runs"",""checkpoint"":""post-trained"",""context_length"":""131,072 / 128K"",""context_management"":""none stated"",""dataset_version"":""MCP-Atlas"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official MCP-Atlas v1, 20-turn limit"",""judge"":""Gemini-2.5-Pro claim coverage; pass at >=0.75"",""max_new_tokens"":""source does not state"",""mode"":""block-diffusion with Levenshtein editing"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""block_length=32; threshold=0.5; editing_threshold=0.0"",""samples"":5,""search"":""none"",""shots"":""source does not state"",""split"":""public 500 tasks"",""temperature"":""1.0"",""timeout"":""source does not state"",""tools"":""36 MCP servers / public tool suite"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":5}",True,official_paper,verified,Displayed exactly as 46.21. Locked official InclusionAI observation.
llada2.2-flash,multi_if,73.67,https://raw.githubusercontent.com/inclusionAI/LLaDA2.X/f402f0b52817e1c3586a024ebf89e69ad8ca5523/LLaDA2_2_tech_report.pdf,"{""aggregation"":""arithmetic mean of five runs"",""checkpoint"":""post-trained"",""context_length"":""131,072 / 128K"",""context_management"":""none stated"",""dataset_version"":""Multi-IF turn-3"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""benchmark-specified"",""max_new_tokens"":""source does not state"",""mode"":""block-diffusion with Levenshtein editing"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""block_length=32; threshold=0.5; editing_threshold=0.0"",""samples"":5,""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""1.0"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":5}",True,official_paper,verified,Displayed exactly as 73.67. Locked official InclusionAI observation.
llada2.2-flash,swe_bench_multilingual,25.0,https://raw.githubusercontent.com/inclusionAI/LLaDA2.X/f402f0b52817e1c3586a024ebf89e69ad8ca5523/LLaDA2_2_tech_report.pdf,"{""aggregation"":""arithmetic mean of five runs"",""checkpoint"":""post-trained"",""context_length"":""131,072 / 128K"",""context_management"":""none stated"",""dataset_version"":""SWE-bench Multilingual"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""Claude Code"",""judge"":""official repository test suite"",""max_new_tokens"":""source does not state"",""mode"":""block-diffusion with Levenshtein editing"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""block_length=32; threshold=0.5; editing_threshold=0.0"",""samples"":5,""search"":""none"",""shots"":""source does not state"",""split"":""Multilingual 300"",""temperature"":""1.0"",""timeout"":""source does not state"",""tools"":""coding-agent shell/editor tools"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":5}",True,official_paper,verified,Displayed exactly as 25.00. Locked official InclusionAI observation.
llada2.2-flash,swe_bench_pro,30.1,https://raw.githubusercontent.com/inclusionAI/LLaDA2.X/f402f0b52817e1c3586a024ebf89e69ad8ca5523/LLaDA2_2_tech_report.pdf,"{""aggregation"":""arithmetic mean of five runs"",""checkpoint"":""post-trained"",""context_length"":""131,072 / 128K"",""context_management"":""none stated"",""dataset_version"":""SWE-bench Pro"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""Claude Code"",""judge"":""official repository test suite"",""max_new_tokens"":""source does not state"",""mode"":""block-diffusion with Levenshtein editing"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""block_length=32; threshold=0.5; editing_threshold=0.0"",""samples"":5,""search"":""none"",""shots"":""source does not state"",""split"":""public 731"",""temperature"":""1.0"",""timeout"":""source does not state"",""tools"":""coding-agent shell/editor tools"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":5}",True,official_paper,verified,Displayed exactly as 30.10. Locked official InclusionAI observation.
llada2.2-flash,swe_bench_verified,49.28,https://raw.githubusercontent.com/inclusionAI/LLaDA2.X/f402f0b52817e1c3586a024ebf89e69ad8ca5523/LLaDA2_2_tech_report.pdf,"{""aggregation"":""arithmetic mean of five runs"",""checkpoint"":""post-trained"",""context_length"":""131,072 / 128K"",""context_management"":""none stated"",""dataset_version"":""SWE-bench Verified"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""OpenHands"",""judge"":""official repository test suite"",""max_new_tokens"":""source does not state"",""mode"":""block-diffusion with Levenshtein editing"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""block_length=32; threshold=0.5; editing_threshold=0.0"",""samples"":5,""search"":""none"",""shots"":""source does not state"",""split"":""Verified 500"",""temperature"":""1.0"",""timeout"":""source does not state"",""tools"":""coding-agent shell/editor tools"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":5}",True,official_paper,verified,Displayed exactly as 49.28. Locked official InclusionAI observation.
llada2.2-flash,tau2_bench_avg,80.33,https://raw.githubusercontent.com/inclusionAI/LLaDA2.X/f402f0b52817e1c3586a024ebf89e69ad8ca5523/LLaDA2_2_tech_report.pdf,"{""aggregation"":""arithmetic mean of five runs"",""checkpoint"":""post-trained"",""context_length"":""131,072 / 128K"",""context_management"":""none stated"",""dataset_version"":""tau2 v1.0.0 average"",""effort"":""default"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official tau2 harness with source-specific user simulator"",""judge"":""task success"",""max_new_tokens"":""source does not state"",""mode"":""block-diffusion with Levenshtein editing"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""block_length=32; threshold=0.5; editing_threshold=0.0"",""samples"":5,""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""1.0"",""timeout"":""source does not state"",""tools"":""official tau2 domain tools"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":5}",True,official_paper,verified,Displayed exactly as 80.33. Locked official InclusionAI observation.
ring-2.6-1t,aa_lcr,64.3,https://intranetproxy.alipay.com/skylark/lark/0/2026/png/23157180/1785831264180-d6ca4404-acef-4424-84db-fbc5a4c6db5f.png,"{""aggregation"":""source does not state"",""checkpoint"":""post-trained"",""context_length"":""131,072 native; card advertises 256K YaRN"",""context_management"":""none stated"",""dataset_version"":""AA-LCR Intelligence Index methodology"",""effort"":""xhigh"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""Artificial Analysis equality-checker LLM"",""max_new_tokens"":""source does not state"",""mode"":""thinking"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":""source does not state"",""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":""source does not state""}",False,official_model_card,verified,Displayed exactly as 64.3. Locked official InclusionAI observation.
ring-2.6-1t,aime_2026,95.83,https://huggingface.co/inclusionAI/Ring-2.6-1T/blob/1e58be9318352541575130d4dbbdcc735fca7a03/.eval_results/aime_2026.yaml,"{""aggregation"":""mean@64"",""checkpoint"":""post-trained"",""context_length"":""131,072 native; card advertises 256K YaRN"",""context_management"":""none stated"",""dataset_version"":""AIME 2026"",""effort"":""xhigh"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""benchmark-specified"",""max_new_tokens"":""source does not state"",""mode"":""thinking"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":64,""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":64}",True,official_model_card,verified,Displayed exactly as 95.83. Locked official InclusionAI observation.
ring-2.6-1t,arc_agi_2,66.18,https://arxiv.org/pdf/2606.15079v1,"{""aggregation"":""pass@2"",""checkpoint"":""post-trained"",""context_length"":""131,072 native; card advertises 256K YaRN"",""context_management"":""none stated"",""dataset_version"":""ARC-AGI-2"",""effort"":""xhigh"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""ARC exact task solver"",""max_new_tokens"":""source does not state"",""mode"":""thinking"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":2,""search"":""none"",""shots"":""source does not state"",""split"":""ARC-AGI-2 public evaluation set"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":2}",True,official_paper,verified,Displayed exactly as 66.18. Locked official InclusionAI observation.
ring-2.6-1t,bfcl_v4,64.8,https://intranetproxy.alipay.com/skylark/lark/0/2026/png/23157180/1785831264180-d6ca4404-acef-4424-84db-fbc5a4c6db5f.png,"{""aggregation"":""source does not state"",""checkpoint"":""post-trained"",""context_length"":""131,072 native; card advertises 256K YaRN"",""context_management"":""none stated"",""dataset_version"":""BFCL v4"",""effort"":""xhigh"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""Berkeley Function Calling Leaderboard"",""judge"":""official function-call match"",""max_new_tokens"":""source does not state"",""mode"":""thinking"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":""source does not state"",""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""benchmark-provided functions"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":""source does not state""}",False,official_model_card,verified,Displayed exactly as 64.8. Locked official InclusionAI observation.
ring-2.6-1t,browsecomp,71.7,https://intranetproxy.alipay.com/skylark/lark/0/2026/png/23157180/1785831264180-d6ca4404-acef-4424-84db-fbc5a4c6db5f.png,"{""aggregation"":""source does not state"",""checkpoint"":""post-trained"",""context_length"":""131,072 native; card advertises 256K YaRN"",""context_management"":""none stated"",""dataset_version"":""BrowseComp"",""effort"":""xhigh"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""InclusionAI internal search-agent harness"",""judge"":""BrowseComp exact-answer judge"",""max_new_tokens"":""source does not state"",""mode"":""thinking"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":""source does not state"",""search"":""enabled"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""browser/search and code execution"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":""source does not state""}",False,official_model_card,verified,Displayed exactly as 71.7. Locked official InclusionAI observation.
ring-2.6-1t,claw_eval_general_pass3,63.82,https://huggingface.co/inclusionAI/Ring-2.6-1T/blob/1e58be9318352541575130d4dbbdcc735fca7a03/.eval_results/claw_eval.yaml,"{""aggregation"":""Pass^3: all three trials must pass"",""checkpoint"":""post-trained"",""context_length"":""131,072 native; card advertises 256K YaRN"",""context_management"":""none stated"",""dataset_version"":""ClawEval"",""effort"":""high"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official Claw-Eval v1.1"",""judge"":""full-trajectory completion/safety/robustness grading"",""max_new_tokens"":""source does not state"",""mode"":""thinking"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":3,""search"":""none"",""shots"":""source does not state"",""split"":""general 161 tasks"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""OpenClaw agent environment"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":3}",True,official_model_card,verified,Displayed exactly as 63.82. Locked official InclusionAI observation.
ring-2.6-1t,gdpval_aa_v2_elo,920.0,https://intranetproxy.alipay.com/skylark/lark/0/2026/png/23157180/1785831264180-d6ca4404-acef-4424-84db-fbc5a4c6db5f.png,"{""aggregation"":""source does not state"",""checkpoint"":""post-trained"",""context_length"":""131,072 native; card advertises 256K YaRN"",""context_management"":""none stated"",""dataset_version"":""GDPval v2-AA"",""effort"":""xhigh"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""GDPval-AA v2; 250 turns; 5-hour timeout"",""judge"":""pairwise judge panel / Elo"",""max_new_tokens"":""source does not state"",""mode"":""thinking"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":""source does not state"",""search"":""none"",""shots"":""source does not state"",""split"":""220 public tasks"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""Stirrup agent tools"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":""source does not state""}",False,official_model_card,verified,Displayed exactly as 920. Locked official InclusionAI observation.
ring-2.6-1t,gpqa_diamond,88.27,https://huggingface.co/inclusionAI/Ring-2.6-1T/blob/1e58be9318352541575130d4dbbdcc735fca7a03/.eval_results/gpqa_diamond.yaml,"{""aggregation"":""pass@1"",""checkpoint"":""post-trained"",""context_length"":""131,072 native; card advertises 256K YaRN"",""context_management"":""none stated"",""dataset_version"":""GPQA-Diamond"",""effort"":""xhigh"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""benchmark-specified"",""max_new_tokens"":""source does not state"",""mode"":""thinking"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":1,""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":1}",True,official_model_card,verified,Displayed exactly as 88.27. Locked official InclusionAI observation.
ring-2.6-1t,hmmt_feb_2026,93.47,https://arxiv.org/pdf/2606.15079v1,"{""aggregation"":""mean@64"",""checkpoint"":""post-trained"",""context_length"":""131,072 native; card advertises 256K YaRN"",""context_management"":""none stated"",""dataset_version"":""HMMT February 2026"",""effort"":""xhigh"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""benchmark-specified"",""max_new_tokens"":""source does not state"",""mode"":""thinking"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":64,""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":64}",True,official_paper,verified,Displayed exactly as 93.47. Locked official InclusionAI observation.
ring-2.6-1t,ifbench,44.6,https://intranetproxy.alipay.com/skylark/lark/0/2026/png/23157180/1785831264180-d6ca4404-acef-4424-84db-fbc5a4c6db5f.png,"{""aggregation"":""source does not state"",""checkpoint"":""post-trained"",""context_length"":""131,072 native; card advertises 256K YaRN"",""context_management"":""none stated"",""dataset_version"":""IFBench"",""effort"":""xhigh"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""benchmark-specified"",""max_new_tokens"":""source does not state"",""mode"":""thinking"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":""source does not state"",""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":""source does not state""}",False,official_model_card,verified,Displayed exactly as 44.6. Locked official InclusionAI observation.
ring-2.6-1t,imo_answerbench,86.12,https://arxiv.org/pdf/2606.15079v1,"{""aggregation"":""mean@8"",""checkpoint"":""post-trained"",""context_length"":""131,072 native; card advertises 256K YaRN"",""context_management"":""none stated"",""dataset_version"":""IMO-AnswerBench"",""effort"":""xhigh"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""benchmark-specified"",""max_new_tokens"":""source does not state"",""mode"":""thinking"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":8,""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":8}",True,official_paper,verified,Displayed exactly as 86.12. Locked official InclusionAI observation.
ring-2.6-1t,livecodebench_v6_2408_2505,86.95,https://arxiv.org/pdf/2606.15079v1,"{""aggregation"":""mean@4"",""checkpoint"":""post-trained"",""context_length"":""131,072 native; card advertises 256K YaRN"",""context_management"":""none stated"",""dataset_version"":""LiveCodeBench v6, 2024-08 to 2025-05"",""effort"":""xhigh"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""official test execution"",""max_new_tokens"":""source does not state"",""mode"":""thinking"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":4,""search"":""none"",""shots"":""source does not state"",""split"":""454 code-generation problems"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""code execution for generated solutions"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":4}",True,official_paper,verified,Displayed exactly as 86.95. Locked official InclusionAI observation.
ring-2.6-1t,mcpatlas,61.2,https://intranetproxy.alipay.com/skylark/lark/0/2026/png/23157180/1785831264180-d6ca4404-acef-4424-84db-fbc5a4c6db5f.png,"{""aggregation"":""source does not state"",""checkpoint"":""post-trained"",""context_length"":""131,072 native; card advertises 256K YaRN"",""context_management"":""none stated"",""dataset_version"":""MCP-Atlas"",""effort"":""xhigh"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official MCP-Atlas v1, 20-turn limit"",""judge"":""Gemini-2.5-Pro claim coverage; pass at >=0.75"",""max_new_tokens"":""source does not state"",""mode"":""thinking"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":""source does not state"",""search"":""none"",""shots"":""source does not state"",""split"":""public 500 tasks"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""36 MCP servers / public tool suite"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":""source does not state""}",False,official_model_card,verified,Displayed exactly as 61.2. Locked official InclusionAI observation.
ring-2.6-1t,multi_if,89.3,https://intranetproxy.alipay.com/skylark/lark/0/2026/png/23157180/1785831264180-d6ca4404-acef-4424-84db-fbc5a4c6db5f.png,"{""aggregation"":""source does not state"",""checkpoint"":""post-trained"",""context_length"":""131,072 native; card advertises 256K YaRN"",""context_management"":""none stated"",""dataset_version"":""Multi-IF turn-3"",""effort"":""xhigh"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official InclusionAI / benchmark-specific"",""judge"":""benchmark-specified"",""max_new_tokens"":""source does not state"",""mode"":""thinking"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":""source does not state"",""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""none"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":""source does not state""}",False,official_model_card,verified,Displayed exactly as 89.3. Locked official InclusionAI observation.
ring-2.6-1t,pinchbench_123,87.6,https://arxiv.org/pdf/2606.15079v1,"{""aggregation"":""mean@3"",""checkpoint"":""post-trained"",""context_length"":""131,072 native; card advertises 256K YaRN"",""context_management"":""none stated"",""dataset_version"":""PinchBench 123-task manifest at commit 27afe091b6ae04ec4b6aa9f5459bd280da0fd61d (2026-04-24)"",""effort"":""high"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official PinchBench/OpenClaw suite"",""judge"":""task-specific automatic and/or LLM judge"",""max_new_tokens"":""source does not state"",""mode"":""thinking"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":3,""search"":""none"",""shots"":""source does not state"",""split"":""123-task manifest at commit 27afe091b6ae04ec4b6aa9f5459bd280da0fd61d (2026-04-24)"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""OpenClaw agent environment"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":3}",True,official_paper,verified,Displayed exactly as 87.60. Locked official InclusionAI observation.
ring-2.6-1t,swe_bench_multilingual,56.7,https://intranetproxy.alipay.com/skylark/lark/0/2026/png/23157180/1785831264180-d6ca4404-acef-4424-84db-fbc5a4c6db5f.png,"{""aggregation"":""source does not state"",""checkpoint"":""post-trained"",""context_length"":""256K"",""context_management"":""none stated"",""dataset_version"":""SWE-bench Multilingual"",""effort"":""xhigh"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""OpenHands with tailored prompts"",""judge"":""official repository test suite"",""max_new_tokens"":""32K"",""mode"":""thinking"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""OpenHands tailored prompts"",""samples"":""source does not state"",""search"":""none"",""shots"":""source does not state"",""split"":""Multilingual 300"",""temperature"":""0.6"",""timeout"":""source does not state"",""tools"":""coding-agent shell/editor tools"",""top_k"":""source does not state"",""top_p"":""0.95"",""trials"":""source does not state""}",False,official_model_card,verified,Displayed exactly as 56.7. Locked official InclusionAI observation. The adjacent benchmark-labeled methodology bullet is scoped row-wide across every displayed model column.
ring-2.6-1t,swe_bench_pro,53.76,https://arxiv.org/pdf/2606.15079v1,"{""aggregation"":""% resolved (pass@1)"",""checkpoint"":""post-trained"",""context_length"":""131,072 native; card advertises 256K YaRN"",""context_management"":""none stated"",""dataset_version"":""SWE-bench Pro"",""effort"":""high"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""Claude Code"",""judge"":""official repository test suite"",""max_new_tokens"":""source does not state"",""mode"":""thinking"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":1,""search"":""none"",""shots"":""source does not state"",""split"":""public 731"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""coding-agent shell/editor tools"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":1}",True,official_paper,verified,Displayed exactly as 53.76. Locked official InclusionAI observation.
ring-2.6-1t,swe_bench_verified,74.0,https://huggingface.co/inclusionAI/Ring-2.6-1T/blob/1e58be9318352541575130d4dbbdcc735fca7a03/.eval_results/swe_bench_verified.yaml,"{""aggregation"":""% resolved (pass@1)"",""checkpoint"":""post-trained"",""context_length"":""131,072 native; card advertises 256K YaRN"",""context_management"":""none stated"",""dataset_version"":""SWE-bench Verified"",""effort"":""high"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""source does not state"",""judge"":""official repository test suite"",""max_new_tokens"":""source does not state"",""mode"":""thinking"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":1,""search"":""none"",""shots"":""source does not state"",""split"":""Verified 500"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""coding-agent shell/editor tools"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":1}",True,official_model_card,verified,Displayed exactly as 74.0. Locked official InclusionAI observation.
ring-2.6-1t,tau2_bench_airline,78.0,https://arxiv.org/pdf/2606.15079v1,"{""aggregation"":""Accuracy"",""checkpoint"":""post-trained"",""context_length"":""131,072 native; card advertises 256K YaRN"",""context_management"":""none stated"",""dataset_version"":""tau2 v1.0.0 airline"",""effort"":""high"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official tau2 harness with source-specific user simulator"",""judge"":""task success"",""max_new_tokens"":""source does not state"",""mode"":""thinking"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":1,""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""official tau2 domain tools"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":1}",True,official_paper,verified,Displayed exactly as 78.00. Locked official InclusionAI observation.
ring-2.6-1t,tau2_bench_avg,84.26,https://arxiv.org/pdf/2606.15079v1,"{""aggregation"":""Accuracy"",""checkpoint"":""post-trained"",""context_length"":""131,072 native; card advertises 256K YaRN"",""context_management"":""none stated"",""dataset_version"":""tau2 v1.0.0 average"",""effort"":""high"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official tau2 harness with source-specific user simulator"",""judge"":""task success"",""max_new_tokens"":""source does not state"",""mode"":""thinking"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":1,""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""official tau2 domain tools"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":1}",True,official_paper,verified,Displayed exactly as 84.26. Locked official InclusionAI observation.
ring-2.6-1t,tau2_bench_retail,78.07,https://arxiv.org/pdf/2606.15079v1,"{""aggregation"":""Accuracy"",""checkpoint"":""post-trained"",""context_length"":""131,072 native; card advertises 256K YaRN"",""context_management"":""none stated"",""dataset_version"":""tau2 v1.0.0 retail"",""effort"":""high"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official tau2 harness with source-specific user simulator"",""judge"":""task success"",""max_new_tokens"":""source does not state"",""mode"":""thinking"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":1,""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""official tau2 domain tools"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":1}",True,official_paper,verified,Displayed exactly as 78.07. Locked official InclusionAI observation.
ring-2.6-1t,tau2_bench_telecom,96.71,https://arxiv.org/pdf/2606.15079v1,"{""aggregation"":""Accuracy"",""checkpoint"":""post-trained"",""context_length"":""131,072 native; card advertises 256K YaRN"",""context_management"":""none stated"",""dataset_version"":""tau2 v1.0.0 telecom"",""effort"":""high"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""official tau2 harness with source-specific user simulator"",""judge"":""task success"",""max_new_tokens"":""source does not state"",""mode"":""thinking"",""multimodal_inputs"":false,""output_parser"":""source does not state"",""prompt_style"":""source does not state"",""samples"":1,""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""source does not state"",""timeout"":""source does not state"",""tools"":""official tau2 domain tools"",""top_k"":""source does not state"",""top_p"":""source does not state"",""trials"":1}",True,official_paper,verified,Displayed exactly as 96.71. Locked official InclusionAI observation.
ring-2.6-1t,terminal_bench_2_1,43.1,https://intranetproxy.alipay.com/skylark/lark/0/2026/png/23157180/1785831264180-d6ca4404-acef-4424-84db-fbc5a4c6db5f.png,"{""aggregation"":""arithmetic mean of 3 runs"",""checkpoint"":""post-trained"",""context_length"":""256K"",""context_management"":""none stated"",""dataset_version"":""Terminal-Bench 2.1, 89 tasks"",""effort"":""xhigh"",""evaluation_provenance"":""official InclusionAI evaluation or locked provider rendering"",""harness"":""Terminus 2 under Artificial Analysis protocol"",""judge"":""programmatic task tests"",""max_new_tokens"":""32K"",""mode"":""thinking"",""multimodal_inputs"":false,""output_parser"":""provided JSON parser in preserve-thinking mode"",""prompt_style"":""source does not state"",""samples"":1,""search"":""none"",""shots"":""source does not state"",""split"":""source does not state"",""temperature"":""0.6"",""timeout"":""2 hours per task"",""tools"":""terminal"",""top_k"":""source does not state"",""top_p"":""1.0"",""trials"":3}",False,official_model_card,verified,Displayed exactly as 43.1. Locked official InclusionAI observation. The adjacent benchmark-labeled methodology bullet is scoped row-wide across every displayed model column.
claude-opus-4.7,deepsearchqa_acc,82.3,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""source-reported score"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""DeepSearchQA 900-prompt eval split"",""effort"":""source does not state"",""harness_agent"":""official StepFun evaluation; exact harness not stated"",""judge"":""Gemini 2.5 Flash with official Kaggle grading prompt"",""mode"":""source does not state"",""multimodal_input"":false,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""web search/research agent"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",False,official_blog,verified,Official StepFun launch comparison table.
claude-opus-4.7,deepsearchqa_f1,91.7,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""source-reported score"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""DeepSearchQA 900-prompt eval split"",""effort"":""source does not state"",""harness_agent"":""official StepFun evaluation; exact harness not stated"",""judge"":""Gemini 2.5 Flash with official Kaggle grading prompt"",""mode"":""source does not state"",""multimodal_input"":false,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""web search/research agent"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",False,official_blog,verified,Official StepFun launch comparison table.
claude-opus-4.7,gdpval_aa_elo,1753.0,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""source-reported score"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""GDPval-AA historical 220-task Stirrup leaderboard quoted 2026-05-29"",""effort"":""source does not state"",""harness_agent"":""official Artificial Analysis Stirrup leaderboard"",""judge"":""pairwise Elo evaluation anchored to human experts"",""mode"":""source does not state"",""multimodal_input"":false,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""official Artificial Analysis Stirrup harness"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",False,official_blog,verified,Official StepFun launch comparison table.
claude-opus-4.7,hle_tools,54.7,https://www-cdn.anthropic.com/images/4zrzovbb/website/d434d15757c6abac1122af483617741776d5a114-2600x2638.png,"{""aggregation_pass_k"":""source-reported score"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""HLE finalized full set"",""effort"":""official release evaluation; exact effort not shown in chart"",""harness_agent"":""official Anthropic HLE with-tools evaluation"",""judge"":""benchmark-specified; exact StepFun judge not stated"",""mode"":""adaptive reasoning"",""multimodal_input"":true,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""tools enabled; exact suite undisclosed"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",False,official_blog,verified,Official StepFun launch comparison table. Direct locked official score authority reports HLE w. tool=54.7; the StepFun surface is retained as physical corroboration.
claude-opus-4.7,researchrubrics,73.9,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""source-reported score"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""ResearchRubrics 101-task official release"",""effort"":""source does not state"",""harness_agent"":""official StepFun evaluation; exact harness not stated"",""judge"":""weighted binary rubric compliance evaluation"",""mode"":""source does not state"",""multimodal_input"":false,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""research tools"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",False,official_blog,verified,Official StepFun launch comparison table.
deepseek-v4-flash,deepsearchqa_acc,79.8,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""source-reported score"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""DeepSearchQA 900-prompt eval split"",""effort"":""source does not state"",""harness_agent"":""official StepFun evaluation; exact harness not stated"",""judge"":""Gemini 2.5 Flash with official Kaggle grading prompt"",""mode"":""source does not state"",""multimodal_input"":false,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""web search/research agent"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",False,official_blog,verified,Official StepFun launch comparison table.
deepseek-v4-flash,hle_tools,45.1,https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro/resolve/b5968e9190ef611bbf34a7229255be88a0e937c1/README.md?download=true,"{""aggregation_pass_k"":""source-reported score"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""HLE finalized full set"",""effort"":""Max"",""harness_agent"":""official DeepSeek HLE tool evaluation"",""judge"":""benchmark-specified; exact StepFun judge not stated"",""mode"":""thinking"",""multimodal_input"":true,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""tools enabled; exact suite undisclosed"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",False,official_model_card,verified,Official StepFun launch comparison table. Direct locked official score authority reports HLE w. tool=45.1; the StepFun surface is retained as physical corroboration.
deepseek-v4-flash,researchrubrics,66.2,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""source-reported score"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""ResearchRubrics 101-task official release"",""effort"":""source does not state"",""harness_agent"":""official StepFun evaluation; exact harness not stated"",""judge"":""weighted binary rubric compliance evaluation"",""mode"":""source does not state"",""multimodal_input"":false,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""research tools"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",False,official_blog,verified,Official StepFun launch comparison table.
deepseek-v4-pro,researchrubrics,68.3,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""source-reported score"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""ResearchRubrics 101-task official release"",""effort"":""source does not state"",""harness_agent"":""official StepFun evaluation; exact harness not stated"",""judge"":""weighted binary rubric compliance evaluation"",""mode"":""source does not state"",""multimodal_input"":false,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""research tools"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",False,official_blog,verified,Official StepFun launch comparison table.
gemini-3-flash,android_daily,63.21,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""pass@1 task success rate"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""AndroidDaily v1, 350 tasks / 94 apps"",""effort"":""source does not state"",""harness_agent"":""StepFun phone-use stack; exact launch harness undisclosed"",""judge"":""benchmark-specified; exact StepFun judge not stated"",""mode"":""source does not state"",""multimodal_input"":true,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""physical Android phone-use action environment"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",False,official_blog,verified,Corrected directly from the current locked official DOM. The prior research's 48.2/43.7/43.3/37.6 transcription was stale and is not retained.
gemini-3.5-flash,researchrubrics,63.6,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""source-reported score"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""ResearchRubrics 101-task official release"",""effort"":""source does not state"",""harness_agent"":""official StepFun evaluation; exact harness not stated"",""judge"":""weighted binary rubric compliance evaluation"",""mode"":""source does not state"",""multimodal_input"":false,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""research tools"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",False,official_blog,verified,Official StepFun launch comparison table.
glm-5.1,deepsearchqa_acc,81.3,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""source-reported score"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""DeepSearchQA 900-prompt eval split"",""effort"":""source does not state"",""harness_agent"":""official StepFun evaluation; exact harness not stated"",""judge"":""Gemini 2.5 Flash with official Kaggle grading prompt"",""mode"":""source does not state"",""multimodal_input"":false,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""web search/research agent"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",False,official_blog,verified,Official StepFun launch comparison table.
glm-5.1,researchrubrics,67.9,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""source-reported score"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""ResearchRubrics 101-task official release"",""effort"":""source does not state"",""harness_agent"":""official StepFun evaluation; exact harness not stated"",""judge"":""weighted binary rubric compliance evaluation"",""mode"":""source does not state"",""multimodal_input"":false,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""research tools"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",False,official_blog,verified,Official StepFun launch comparison table.
glm-5v-turbo,android_daily,51.68,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""pass@1 task success rate"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""AndroidDaily v1, 350 tasks / 94 apps"",""effort"":""source does not state"",""harness_agent"":""StepFun phone-use stack; exact launch harness undisclosed"",""judge"":""benchmark-specified; exact StepFun judge not stated"",""mode"":""source does not state"",""multimodal_input"":true,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""physical Android phone-use action environment"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",False,official_blog,verified,Corrected directly from the current locked official DOM. The prior research's 48.2/43.7/43.3/37.6 transcription was stale and is not retained.
glm-5v-turbo,hr_bench_4k,84.62,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""accuracy"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""HR-Bench 4K official 800-row release"",""effort"":""source does not state"",""harness_agent"":""provider-aligned crop + search and other visual tools; exact tool composition not disclosed"",""judge"":""benchmark-specified; exact StepFun judge not stated"",""mode"":""source does not state"",""multimodal_input"":true,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""Python visual tool: crop/zoom/draw/search actions"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",False,official_blog,verified,Official StepFun benchmark-specific tool table.
glm-5v-turbo,hr_bench_8k,83.12,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""accuracy"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""HR-Bench 8K official 800-row release"",""effort"":""source does not state"",""harness_agent"":""provider-aligned crop + search and other visual tools; exact tool composition not disclosed"",""judge"":""benchmark-specified; exact StepFun judge not stated"",""mode"":""source does not state"",""multimodal_input"":true,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""Python visual tool: crop/zoom/draw/search actions"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",False,official_blog,verified,Official StepFun benchmark-specific tool table.
glm-5v-turbo,simplevqa,78.2,https://cdn.bigmodel.cn/markdown/1775044860800img_v3_0210b_c42d0355-22af-4be7-8582-16fc6e7d0bfg.png?attname=img_v3_0210b_c42d0355-22af-4be7-8582-16fc6e7d0bfg.png,"{""aggregation_pass_k"":""source-reported score"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""SimpleVQA official 2,025-item test release"",""effort"":""source does not state"",""harness_agent"":""official GLM multimodal ToolUse evaluation; exact tool composition not disclosed"",""judge"":""LLM-as-judge; exact StepFun judge not stated"",""mode"":""source does not state"",""multimodal_input"":true,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""Visual Search"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",False,official_blog,verified,Official StepFun benchmark-specific tool table. Direct locked official score authority reports SimpleVQA=78.2; the StepFun surface is retained as physical corroboration.
glm-5v-turbo,vstar,89.0,https://cdn.bigmodel.cn/markdown/1775044860800img_v3_0210b_c42d0355-22af-4be7-8582-16fc6e7d0bfg.png?attname=img_v3_0210b_c42d0355-22af-4be7-8582-16fc6e7d0bfg.png,"{""aggregation_pass_k"":""accuracy"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""V* Bench official 191-question release"",""effort"":""source does not state"",""harness_agent"":""official GLM multimodal ToolUse evaluation; exact tool composition not disclosed"",""judge"":""benchmark-specified; exact StepFun judge not stated"",""mode"":""source does not state"",""multimodal_input"":true,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""Python visual tool: crop/zoom/draw/search actions"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",False,official_blog,verified,Official StepFun benchmark-specific tool table. Direct locked official score authority reports V*=89.0; the StepFun surface is retained as physical corroboration.
glm-5v-turbo,worldvqa,47.81,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""source-reported score"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""WorldVQA first-eight-category 3,000-item release"",""effort"":""source does not state"",""harness_agent"":""official StepFun evaluation; exact harness not stated"",""judge"":""LLM-as-judge; exact StepFun judge not stated"",""mode"":""source does not state"",""multimodal_input"":true,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""Visual Search"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",False,official_blog,verified,Official StepFun benchmark-specific tool table.
gpt-5.4,claw_eval_pass3,60.3,https://raw.githubusercontent.com/claw-eval/claw-eval.github.io/e10516b5c2a6dfc139be35dacb54cdf5e034801d/assets/benchmark-BgnHRA8t.js,"{""aggregation_pass_k"":""Pass^3: all three trials must pass"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""Claw-Eval v1.1 non-multimodal: 161 general + 38 multi-turn tasks"",""effort"":""source does not state"",""harness_agent"":""official Claw-Eval v1.1 non-multimodal leaderboard"",""judge"":""full-trajectory completion/safety/robustness grading"",""mode"":""source does not state"",""multimodal_input"":false,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""official Claw-Eval agent environment"",""top_p"":""source does not state"",""trials_samples"":3}",True,official_repository,verified,Official launch hero chart label. Direct locked official score authority reports Claweval-v1.1=60.3; the StepFun surface is retained as physical corroboration.
kimi-k2.6,android_daily,53.36,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""pass@1 task success rate"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""AndroidDaily v1, 350 tasks / 94 apps"",""effort"":""source does not state"",""harness_agent"":""StepFun phone-use stack; exact launch harness undisclosed"",""judge"":""benchmark-specified; exact StepFun judge not stated"",""mode"":""source does not state"",""multimodal_input"":true,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""physical Android phone-use action environment"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",False,official_blog,verified,Corrected directly from the current locked official DOM. The prior research's 48.2/43.7/43.3/37.6 transcription was stale and is not retained.
kimi-k2.6,hr_bench_4k,91.25,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""accuracy"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""HR-Bench 4K official 800-row release"",""effort"":""source does not state"",""harness_agent"":""official StepFun evaluation; exact harness not stated"",""judge"":""benchmark-specified; exact StepFun judge not stated"",""mode"":""source does not state"",""multimodal_input"":true,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""Python visual tool: crop/zoom/draw/search actions"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",False,official_blog,verified,Official StepFun benchmark-specific tool table.
kimi-k2.6,hr_bench_8k,90.13,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""accuracy"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""HR-Bench 8K official 800-row release"",""effort"":""source does not state"",""harness_agent"":""official StepFun evaluation; exact harness not stated"",""judge"":""benchmark-specified; exact StepFun judge not stated"",""mode"":""source does not state"",""multimodal_input"":true,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""Python visual tool: crop/zoom/draw/search actions"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",False,official_blog,verified,Official StepFun benchmark-specific tool table.
kimi-k2.6,researchrubrics,63.0,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""source-reported score"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""ResearchRubrics 101-task official release"",""effort"":""source does not state"",""harness_agent"":""official StepFun evaluation; exact harness not stated"",""judge"":""weighted binary rubric compliance evaluation"",""mode"":""source does not state"",""multimodal_input"":false,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""research tools"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",False,official_blog,verified,Official StepFun launch comparison table.
kimi-k2.6,simplevqa,78.24,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""source-reported score"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""SimpleVQA official 2,025-item test release"",""effort"":""source does not state"",""harness_agent"":""official StepFun evaluation; exact harness not stated"",""judge"":""LLM-as-judge; exact StepFun judge not stated"",""mode"":""source does not state"",""multimodal_input"":true,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""Visual Search"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",False,official_blog,verified,Official StepFun benchmark-specific tool table.
kimi-k2.6,vstar,96.9,https://huggingface.co/moonshotai/Kimi-K2.6/resolve/d9cb81bc88b9bd2dc89877599f23c614ded72b9c/README.md?download=true,"{""aggregation_pass_k"":""accuracy"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""V* Bench official 191-question release"",""effort"":""source does not state"",""harness_agent"":""official Kimi V* evaluation with Python visual tool"",""judge"":""benchmark-specified; exact StepFun judge not stated"",""mode"":""source does not state"",""multimodal_input"":true,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""Python visual tool: crop/zoom/draw/search actions"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",False,official_model_card,verified,Official StepFun benchmark-specific tool table. Direct locked official score authority reports V*=96.9; the StepFun surface is retained as physical corroboration.
kimi-k2.6,worldvqa,55.98,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""source-reported score"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""WorldVQA first-eight-category 3,000-item release"",""effort"":""source does not state"",""harness_agent"":""official StepFun evaluation; exact harness not stated"",""judge"":""LLM-as-judge; exact StepFun judge not stated"",""mode"":""source does not state"",""multimodal_input"":true,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""Visual Search"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",False,official_blog,verified,Official StepFun benchmark-specific tool table.
step-3.5-flash,aa_lcr,45.5,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""average accuracy over 16 samples"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""AA-LCR 100-question release"",""effort"":""source does not state"",""harness_agent"":""official StepFun evaluation; exact harness not stated"",""judge"":""AA equality checker; StepFun exact judge revision undisclosed"",""mode"":""source does not state"",""multimodal_input"":false,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""none"",""top_p"":""source does not state"",""trials_samples"":16}",False,official_blog,verified,Official StepFun launch comparison table.
step-3.5-flash,browsecomp,69.0,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""source-reported score"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""BrowseComp 1,266-question release"",""effort"":""source does not state"",""harness_agent"":""official StepFun evaluation; exact harness not stated"",""judge"":""benchmark-specified; exact StepFun judge not stated"",""mode"":""source does not state"",""multimodal_input"":false,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""web browsing agent"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",True,official_blog,verified,Official StepFun launch comparison table.
step-3.5-flash,claw_eval_pass3,43.6,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""Pass^3: all three trials must pass"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""Claw-Eval v1.1 non-multimodal: 161 general + 38 multi-turn tasks"",""effort"":""source does not state"",""harness_agent"":""official StepFun evaluation; exact harness not stated"",""judge"":""full-trajectory completion/safety/robustness grading"",""mode"":""source does not state"",""multimodal_input"":false,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""official Claw-Eval agent environment"",""top_p"":""source does not state"",""trials_samples"":3}",True,official_blog,verified,Official StepFun launch comparison table.
step-3.5-flash,deepsearchqa_acc,73.4,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""source-reported score"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""DeepSearchQA 900-prompt eval split"",""effort"":""source does not state"",""harness_agent"":""official StepFun evaluation; exact harness not stated"",""judge"":""Gemini 2.5 Flash with official Kaggle grading prompt"",""mode"":""source does not state"",""multimodal_input"":false,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""web search/research agent"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",True,official_blog,verified,Official StepFun launch comparison table.
step-3.5-flash,deepsearchqa_f1,85.5,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""source-reported score"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""DeepSearchQA 900-prompt eval split"",""effort"":""source does not state"",""harness_agent"":""official StepFun evaluation; exact harness not stated"",""judge"":""Gemini 2.5 Flash with official Kaggle grading prompt"",""mode"":""source does not state"",""multimodal_input"":false,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""web search/research agent"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",True,official_blog,verified,Official StepFun launch comparison table.
step-3.5-flash,gdpval_aa_elo,1055.0,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""source-reported score"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""GDPval-AA historical 220-task Stirrup leaderboard quoted 2026-05-29"",""effort"":""source does not state"",""harness_agent"":""official Artificial Analysis Stirrup leaderboard"",""judge"":""pairwise Elo evaluation anchored to human experts"",""mode"":""source does not state"",""multimodal_input"":false,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""official Artificial Analysis Stirrup harness"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",False,official_blog,verified,Official StepFun launch comparison table.
step-3.5-flash,hle_tools,35.7,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""source-reported score"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""HLE finalized full set"",""effort"":""source does not state"",""harness_agent"":""official StepFun evaluation; exact harness not stated"",""judge"":""benchmark-specified; exact StepFun judge not stated"",""mode"":""source does not state"",""multimodal_input"":true,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""tools enabled; exact suite undisclosed"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",True,official_blog,verified,Official StepFun launch comparison table.
step-3.5-flash,researchrubrics,65.3,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""source-reported score"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""ResearchRubrics 101-task official release"",""effort"":""source does not state"",""harness_agent"":""official StepFun evaluation; exact harness not stated"",""judge"":""weighted binary rubric compliance evaluation"",""mode"":""source does not state"",""multimodal_input"":false,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""research tools"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",True,official_blog,verified,Official StepFun launch comparison table.
step-3.5-flash,swe_bench_multilingual,67.4,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""% resolved"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""SWE-bench Multilingual official 300-instance test set"",""effort"":""source does not state"",""harness_agent"":""source does not state coding scaffold"",""judge"":""benchmark-specified; exact StepFun judge not stated"",""mode"":""source does not state"",""multimodal_input"":false,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""agentic coding tools"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",True,official_blog,verified,Official StepFun launch comparison table.
step-3.5-flash,swe_bench_pro,51.3,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""% resolved"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""source label only"",""effort"":""source does not state"",""harness_agent"":""source does not state coding scaffold"",""judge"":""benchmark-specified; exact StepFun judge not stated"",""mode"":""source does not state"",""multimodal_input"":false,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""agentic coding tools"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",True,official_blog,verified,Official StepFun launch comparison table.
step-3.5-flash,terminal_bench_2_1,53.4,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""% tasks resolved"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""Terminal-Bench 2.1"",""effort"":""source does not state"",""harness_agent"":""source does not state terminal scaffold"",""judge"":""benchmark-specified; exact StepFun judge not stated"",""mode"":""source does not state"",""multimodal_input"":false,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""terminal tools"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",False,official_blog,verified,Official StepFun launch comparison table.
step-3.7-flash,aa_lcr,63.9,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""average accuracy over 16 samples"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""AA-LCR 100-question release"",""effort"":""source does not state"",""harness_agent"":""official StepFun evaluation; exact harness not stated"",""judge"":""AA equality checker; StepFun exact judge revision undisclosed"",""mode"":""source does not state"",""multimodal_input"":false,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""none"",""top_p"":""source does not state"",""trials_samples"":16}",False,official_blog,verified,Official StepFun launch comparison table.
step-3.7-flash,android_daily,61.87,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""pass@1 task success rate"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""AndroidDaily v1, 350 tasks / 94 apps"",""effort"":""source does not state"",""harness_agent"":""StepFun phone-use stack; exact launch harness undisclosed"",""judge"":""benchmark-specified; exact StepFun judge not stated"",""mode"":""source does not state"",""multimodal_input"":true,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""physical Android phone-use action environment"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",True,official_blog,verified,Corrected directly from the current locked official DOM. The prior research's 48.2/43.7/43.3/37.6 transcription was stale and is not retained.
step-3.7-flash,browsecomp,75.8,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""source-reported score"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""BrowseComp 1,266-question release"",""effort"":""source does not state"",""harness_agent"":""official StepFun evaluation; exact harness not stated"",""judge"":""benchmark-specified; exact StepFun judge not stated"",""mode"":""source does not state"",""multimodal_input"":false,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""web browsing agent"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",True,official_blog,verified,Official StepFun launch comparison table.
step-3.7-flash,claw_eval_pass3,67.1,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""Pass^3: all three trials must pass"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""Claw-Eval v1.1 non-multimodal: 161 general + 38 multi-turn tasks"",""effort"":""source does not state"",""harness_agent"":""official StepFun evaluation; exact harness not stated"",""judge"":""full-trajectory completion/safety/robustness grading"",""mode"":""source does not state"",""multimodal_input"":false,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""official Claw-Eval agent environment"",""top_p"":""source does not state"",""trials_samples"":3}",True,official_blog,verified,Official StepFun launch comparison table.
step-3.7-flash,deepsearchqa_acc,81.7,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""source-reported score"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""DeepSearchQA 900-prompt eval split"",""effort"":""source does not state"",""harness_agent"":""official StepFun evaluation; exact harness not stated"",""judge"":""Gemini 2.5 Flash with official Kaggle grading prompt"",""mode"":""source does not state"",""multimodal_input"":false,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""web search/research agent"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",True,official_blog,verified,Official StepFun launch comparison table.
step-3.7-flash,deepsearchqa_f1,92.8,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""source-reported score"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""DeepSearchQA 900-prompt eval split"",""effort"":""source does not state"",""harness_agent"":""official StepFun evaluation; exact harness not stated"",""judge"":""Gemini 2.5 Flash with official Kaggle grading prompt"",""mode"":""source does not state"",""multimodal_input"":false,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""web search/research agent"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",True,official_blog,verified,Official StepFun launch comparison table.
step-3.7-flash,gpqa_diamond,78.41,https://raw.githubusercontent.com/stepfun-ai/Step-3.7-Flash/e2132efea2836f5d1638f8145e2178b2f6c9c747/README.md,"{""aggregation_pass_k"":""item accuracy over 3,168 records"",""configuration"":""NVFP4 base decoding without MTP speculative decoding"",""context"":""source does not state"",""dataset_version_split"":""GPQA Diamond 198 items"",""effort"":""source does not state"",""harness_agent"":""official StepFun evaluation; exact harness not stated"",""judge"":""benchmark-specified; exact StepFun judge not stated"",""mode"":""source does not state"",""multimodal_input"":false,""output_cap"":""source does not state"",""quantization_config"":""NVFP4 weights/activations with FP8 KV cache"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""none"",""top_p"":""source does not state"",""trials_samples"":16}",False,official_repository,verified,"Configuration-specific quality result over 3,168 records; kept distinct from non-quantized/common evaluation."
step-3.7-flash,hle_text,49.7,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""source-reported score"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""HLE finalized 2,158-question text-only subset"",""effort"":""source does not state"",""harness_agent"":""official StepFun evaluation; exact harness not stated"",""judge"":""benchmark-specified; exact StepFun judge not stated"",""mode"":""source does not state"",""multimodal_input"":false,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""none"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",True,official_blog,verified,Physically embedded alternative in the HLE tools cell; mapped to the distinct text-only identity.
step-3.7-flash,hle_tools,48.1,https://raw.githubusercontent.com/stepfun-ai/Step-3.7-Flash/e2132efea2836f5d1638f8145e2178b2f6c9c747/README.md,"{""aggregation_pass_k"":""source-reported score"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""HLE finalized full set"",""effort"":""source does not state"",""harness_agent"":""official StepFun evaluation; exact harness not stated"",""judge"":""benchmark-specified; exact StepFun judge not stated"",""mode"":""source does not state"",""multimodal_input"":true,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""tools enabled; exact suite undisclosed"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",True,official_repository,verified,"The post-launch repository prose reports 48.1 tools, distinct from the 47.2 launch table/chart."
step-3.7-flash,hr_bench_4k,89.13,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""accuracy"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""HR-Bench 4K official 800-row release"",""effort"":""source does not state"",""harness_agent"":""official StepFun evaluation; exact harness not stated"",""judge"":""benchmark-specified; exact StepFun judge not stated"",""mode"":""source does not state"",""multimodal_input"":true,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""Python visual tool: crop/zoom/draw/search actions"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",False,official_blog,verified,Official StepFun benchmark-specific tool table.
step-3.7-flash,hr_bench_8k,86.34,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""accuracy"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""HR-Bench 8K official 800-row release"",""effort"":""source does not state"",""harness_agent"":""official StepFun evaluation; exact harness not stated"",""judge"":""benchmark-specified; exact StepFun judge not stated"",""mode"":""source does not state"",""multimodal_input"":true,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""Python visual tool: crop/zoom/draw/search actions"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",False,official_blog,verified,Official StepFun benchmark-specific tool table.
step-3.7-flash,researchrubrics,71.7,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""source-reported score"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""ResearchRubrics 101-task official release"",""effort"":""source does not state"",""harness_agent"":""official StepFun evaluation; exact harness not stated"",""judge"":""weighted binary rubric compliance evaluation"",""mode"":""source does not state"",""multimodal_input"":false,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""research tools"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",True,official_blog,verified,Official StepFun launch comparison table.
step-3.7-flash,simplevqa,79.16,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""source-reported score"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""SimpleVQA official 2,025-item test release"",""effort"":""source does not state"",""harness_agent"":""official StepFun evaluation; exact harness not stated"",""judge"":""LLM-as-judge; exact StepFun judge not stated"",""mode"":""source does not state"",""multimodal_input"":true,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""Visual Search"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",False,official_blog,verified,Official StepFun benchmark-specific tool table.
step-3.7-flash,swe_bench_multilingual,72.4,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""% resolved"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""SWE-bench Multilingual official 300-instance test set"",""effort"":""source does not state"",""harness_agent"":""source does not state coding scaffold"",""judge"":""benchmark-specified; exact StepFun judge not stated"",""mode"":""source does not state"",""multimodal_input"":false,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""agentic coding tools"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",True,official_blog,verified,Official StepFun launch comparison table.
step-3.7-flash,swe_bench_pro,56.3,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""% resolved"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""source label only"",""effort"":""source does not state"",""harness_agent"":""source does not state coding scaffold"",""judge"":""benchmark-specified; exact StepFun judge not stated"",""mode"":""source does not state"",""multimodal_input"":false,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""agentic coding tools"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",True,official_blog,verified,Official StepFun launch comparison table.
step-3.7-flash,swe_bench_verified,76.5,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""% resolved"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""source label only"",""effort"":""source does not state"",""harness_agent"":""source does not state coding scaffold"",""judge"":""benchmark-specified; exact StepFun judge not stated"",""mode"":""source does not state"",""multimodal_input"":false,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""agentic coding tools"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",True,official_blog,verified,Official StepFun launch comparison table.
step-3.7-flash,terminal_bench_2_1,59.55,https://static.stepfun.com/blog/step-3.7-flash/script.js,"{""aggregation_pass_k"":""% tasks resolved"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""Terminal-Bench 2.1"",""effort"":""source does not state"",""harness_agent"":""source does not state terminal scaffold"",""judge"":""benchmark-specified; exact StepFun judge not stated"",""mode"":""source does not state"",""multimodal_input"":false,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""terminal tools"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",False,official_blog,verified,Official StepFun launch comparison table. Physical rendering 59.6 is a rounded display of the locked official script value 59.55; semantic mutation uses the precise value and this rendering is a duplicate. Direct locked official score authority reports Terminal-Bench 2.1=59.55; the StepFun surface is retained as physical corroboration.
step-3.7-flash,vstar,95.29,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""accuracy"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""V* Bench official 191-question release"",""effort"":""source does not state"",""harness_agent"":""official StepFun evaluation; exact harness not stated"",""judge"":""benchmark-specified; exact StepFun judge not stated"",""mode"":""source does not state"",""multimodal_input"":true,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""Python visual tool: crop/zoom/draw/search actions"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",False,official_blog,verified,Official StepFun benchmark-specific tool table.
step-3.7-flash,worldvqa,58.1,https://static.stepfun.com/blog/step-3.7-flash/,"{""aggregation_pass_k"":""source-reported score"",""configuration"":""base Step evaluation configuration"",""context"":""source does not state"",""dataset_version_split"":""WorldVQA first-eight-category 3,000-item release"",""effort"":""source does not state"",""harness_agent"":""official StepFun evaluation; exact harness not stated"",""judge"":""LLM-as-judge; exact StepFun judge not stated"",""mode"":""source does not state"",""multimodal_input"":true,""output_cap"":""source does not state"",""quantization_config"":""common release evaluation; artifact precision not established"",""shots"":""source does not state"",""temperature"":""source does not state"",""tools_search"":""Visual Search"",""top_p"":""source does not state"",""trials_samples"":""source does not state""}",False,official_blog,verified,Official StepFun benchmark-specific tool table.
claude-sonnet-5,code_arena_frontend_elo,1540.0,https://arena.ai/leaderboard/code,"{""aggregation"":""Arena Elo from pairwise human votes"",""context"":""not disclosed"",""dataset_split"":""live WebDev Arena"",""dataset_version"":""Code Arena WebDev live leaderboard"",""effort"":""high"",""harness"":""Code Arena WebDev"",""judge"":""human pairwise preference votes"",""mode"":""thinking"",""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""not disclosed"",""temperature"":""not disclosed"",""tools"":""web application generation"",""trials"":""not disclosed""}",True,leaderboard,verified,Current official leaderboard update; the Google launch table remains preserved as the August release snapshot.
claude-sonnet-5,frontiercode_main_v1_1,42.7,https://storage.googleapis.com/deepmind-media/gemini/gemini_3-7_flash_model_evaluation.pdf,"{""context"":""not disclosed"",""dataset_split"":""100-task Main split"",""dataset_version"":""FrontierCode 1.1"",""effort"":""max"",""harness"":""FrontierCode 1.1 official leaderboard"",""judge"":""blocking held-out functional criteria plus weighted maintainer-authored rubric criteria"",""mode"":""thinking"",""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""not disclosed"",""temperature"":""not disclosed"",""tools"":""coding agent with guarded internet access"",""trials"":""not disclosed""}",False,tech_report,verified,Score; production code quality. Physical rendering retained even when semantically identical to another official rendering.
claude-sonnet-5,gdp_pdf,28.0,https://storage.googleapis.com/deepmind-media/gemini/gemini_3-7_flash_model_evaluation.pdf,"{""aggregation"":""strict all_pass/mean"",""context"":""not disclosed"",""dataset_split"":""test"",""dataset_version"":""GDP.pdf 100-task held-out test set"",""effort"":""max"",""harness"":""Google self-computed GDP.pdf evaluation"",""judge"":""Gemini 3.5 Flash rubric judge"",""mode"":""thinking"",""multimodal_input"":true,""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""not disclosed"",""temperature"":""not disclosed"",""tools"":""none"",""trials"":""not disclosed""}",False,tech_report,verified,Strict task pass rate; expert PDF comprehension. Physical rendering retained even when semantically identical to another official rendering.
claude-sonnet-5,harvey_lab_aa,90.1,https://storage.googleapis.com/deepmind-media/gemini/gemini_3-7_flash_model_evaluation.pdf,"{""aggregation"":""rubric-criterion pass rate"",""context"":""not disclosed"",""dataset_split"":""120 tasks"",""dataset_version"":""Harvey LAB-AA"",""effort"":""max"",""harness"":""Artificial Analysis Harvey LAB-AA"",""judge"":""rubric criteria over extracted deliverables"",""mode"":""thinking"",""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""not disclosed"",""temperature"":""not disclosed"",""tools"":""Stirrup shell and document workflow tools"",""trials"":""not disclosed""}",False,tech_report,verified,Complex legal workflows. Physical rendering retained even when semantically identical to another official rendering.
claude-sonnet-5,hle_verified,31.0,https://storage.googleapis.com/deepmind-media/gemini/gemini_3-7_flash_model_evaluation.pdf,"{""context"":""not disclosed"",""dataset_split"":""Gold 668 plus Revision 1,143"",""dataset_version"":""HLE-Verified full verified set"",""effort"":""max"",""harness"":""Google self-computed HLE-Verified evaluation"",""judge"":""HLE-Verified answer evaluator"",""mode"":""thinking"",""multimodal_input"":true,""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""not disclosed"",""temperature"":""not disclosed"",""tools"":""none"",""trials"":""not disclosed""}",False,tech_report,verified,"1,811 items; Gold + Revision; Uncertain excluded. Physical rendering retained even when semantically identical to another official rendering."
claude-sonnet-5,labbench2,80.1,https://storage.googleapis.com/deepmind-media/gemini/gemini_3-7_flash_model_evaluation.pdf,"{""context"":""not disclosed"",""dataset_split"":""full 1,912-task suite"",""dataset_version"":""LABBench2"",""effort"":""max"",""harness"":""provider self-report; exact harness not disclosed"",""judge"":""not disclosed"",""mode"":""thinking"",""multimodal_input"":true,""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""not disclosed"",""temperature"":""not disclosed"",""tools"":""not disclosed"",""trials"":""not disclosed""}",False,tech_report,verified,Biology real-world research tasks. Physical rendering retained even when semantically identical to another official rendering.
claude-sonnet-5,lvbench,68.5,https://storage.googleapis.com/deepmind-media/gemini/gemini_3-7_flash_model_evaluation.pdf,"{""context"":""not disclosed"",""dataset_split"":""103 videos; 1,549 questions"",""dataset_version"":""LVBench"",""effort"":""max"",""harness"":""Google self-computed LVBench evaluation"",""judge"":""exact multiple-choice answer"",""mode"":""thinking"",""multimodal_input"":true,""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""not disclosed"",""temperature"":""not disclosed"",""tools"":""none"",""trials"":""not disclosed"",""video_frames"":300}",False,tech_report,verified,Long video understanding. Physical rendering retained even when semantically identical to another official rendering.
gemini-3.6-flash,code_arena_frontend_elo,1538.0,https://storage.googleapis.com/deepmind-media/gemini/gemini_3-7_flash_model_evaluation.pdf,"{""aggregation"":""Arena Elo from pairwise human votes"",""context"":""not disclosed"",""dataset_split"":""live WebDev Arena"",""dataset_version"":""Code Arena WebDev live leaderboard"",""effort"":""not disclosed"",""harness"":""Code Arena WebDev"",""judge"":""human pairwise preference votes"",""mode"":""thinking"",""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""not disclosed"",""temperature"":""not disclosed"",""tools"":""web application generation"",""trials"":""not disclosed""}",True,tech_report,verified,Elo. Physical rendering retained even when semantically identical to another official rendering.
gemini-3.6-flash,frontiercode_main_v1_1,34.4,https://storage.googleapis.com/deepmind-media/gemini/gemini_3-7_flash_model_evaluation.pdf,"{""context"":""not disclosed"",""dataset_split"":""100-task Main split"",""dataset_version"":""FrontierCode 1.1"",""effort"":""not disclosed"",""harness"":""FrontierCode 1.1 official leaderboard"",""judge"":""blocking held-out functional criteria plus weighted maintainer-authored rubric criteria"",""mode"":""thinking"",""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""not disclosed"",""temperature"":""not disclosed"",""tools"":""coding agent with guarded internet access"",""trials"":""not disclosed""}",True,tech_report,verified,Score; production code quality. Physical rendering retained even when semantically identical to another official rendering.
gemini-3.6-flash,gdp_pdf,22.0,https://storage.googleapis.com/deepmind-media/gemini/gemini_3-7_flash_model_evaluation.pdf,"{""aggregation"":""strict all_pass/mean"",""context"":""not disclosed"",""dataset_split"":""test"",""dataset_version"":""GDP.pdf 100-task held-out test set"",""effort"":""not disclosed"",""harness"":""Google self-computed GDP.pdf evaluation"",""judge"":""Gemini 3.5 Flash rubric judge"",""mode"":""thinking"",""multimodal_input"":true,""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""pass@1"",""temperature"":""not disclosed"",""tools"":""none"",""trials"":""not disclosed""}",True,tech_report,verified,Strict task pass rate; expert PDF comprehension. Physical rendering retained even when semantically identical to another official rendering.
gemini-3.6-flash,harvey_lab_aa,85.1,https://storage.googleapis.com/deepmind-media/gemini/gemini_3-7_flash_model_evaluation.pdf,"{""aggregation"":""rubric-criterion pass rate"",""context"":""not disclosed"",""dataset_split"":""120 tasks"",""dataset_version"":""Harvey LAB-AA"",""effort"":""not disclosed"",""harness"":""Artificial Analysis Harvey LAB-AA"",""judge"":""rubric criteria over extracted deliverables"",""mode"":""thinking"",""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""not disclosed"",""temperature"":""not disclosed"",""tools"":""Stirrup shell and document workflow tools"",""trials"":""not disclosed""}",True,tech_report,verified,Complex legal workflows. Physical rendering retained even when semantically identical to another official rendering.
gemini-3.6-flash,hle_verified,51.2,https://storage.googleapis.com/deepmind-media/gemini/gemini_3-7_flash_model_evaluation.pdf,"{""context"":""not disclosed"",""dataset_split"":""Gold 668 plus Revision 1,143"",""dataset_version"":""HLE-Verified full verified set"",""effort"":""not disclosed"",""harness"":""Google self-computed HLE-Verified evaluation"",""judge"":""HLE-Verified answer evaluator"",""mode"":""thinking"",""multimodal_input"":true,""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""pass@1"",""temperature"":""not disclosed"",""tools"":""none"",""trials"":""not disclosed""}",True,tech_report,verified,"1,811 items; Gold + Revision; Uncertain excluded. Physical rendering retained even when semantically identical to another official rendering."
gemini-3.6-flash,labbench2,76.1,https://storage.googleapis.com/deepmind-media/gemini/gemini_3-7_flash_model_evaluation.pdf,"{""context"":""not disclosed"",""dataset_split"":""full 1,912-task suite"",""dataset_version"":""LABBench2"",""effort"":""not disclosed"",""harness"":""Google self-computed LABBench2 evaluation"",""judge"":""task-specific exact/programmatic evaluator"",""mode"":""thinking"",""multimodal_input"":true,""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""pass@1"",""temperature"":""not disclosed"",""tools"":""Linux terminal, bioinformatics tools, Python, R, and internet"",""trials"":""not disclosed""}",True,tech_report,verified,Biology real-world research tasks. Physical rendering retained even when semantically identical to another official rendering.
gemini-3.6-flash,lvbench,84.2,https://storage.googleapis.com/deepmind-media/gemini/gemini_3-7_flash_model_evaluation.pdf,"{""context"":""not disclosed"",""dataset_split"":""103 videos; 1,549 questions"",""dataset_version"":""LVBench"",""effort"":""not disclosed"",""harness"":""Google self-computed LVBench evaluation"",""judge"":""exact multiple-choice answer"",""mode"":""thinking"",""multimodal_input"":true,""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""pass@1"",""temperature"":""not disclosed"",""tools"":""none"",""trials"":""not disclosed"",""video_frames"":1024}",True,tech_report,verified,Long video understanding. Physical rendering retained even when semantically identical to another official rendering.
gemini-3.6-flash,osworld_2_0_2026_06_24_partial,33.8,https://storage.googleapis.com/deepmind-media/gemini/gemini_3-7_flash_model_evaluation.pdf,"{""aggregation"":""maximum partial score over three single-attempt runs"",""context"":""not disclosed"",""dataset_split"":""108 tasks; before 2026-08-08 patch"",""dataset_version"":""osworld-v2-2026.06.24"",""effort"":""not disclosed"",""harness"":""OSWorld 2.0 Docker with official Gemini/Anthropic CUA harness"",""judge"":""official weighted checkpoint partial scorer"",""max_steps"":500,""mode"":""thinking"",""multimodal_input"":true,""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""pass@1"",""screen"":""1080p screenshot-only observation"",""temperature"":""not disclosed"",""tools"":""pyautogui actuation plus UI-specific function declarations"",""trials"":3}",True,tech_report,verified,Weighted checkpoint partial score. Physical rendering retained even when semantically identical to another official rendering.
gemini-3.6-flash,terminal_bench_3_0,5.4,https://storage.googleapis.com/deepmind-media/gemini/gemini_3-7_flash_model_evaluation.pdf,"{""context"":""not disclosed"",""dataset_split"":""74 tasks"",""dataset_version"":""Terminal-Bench 3.0 / FrontierBench"",""effort"":""not disclosed"",""harness"":""Google self-computed mini-swe-agent with LiteLLM 1.96; two documented task compatibility modifications"",""judge"":""Terminal-Bench 3 task-success evaluator"",""mode"":""thinking"",""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""pass@1"",""temperature"":""not disclosed"",""tools"":""container terminal"",""trials"":""not disclosed""}",True,tech_report,verified,General agent capabilities. Physical rendering retained even when semantically identical to another official rendering.
gemini-3.7-flash,charxiv_reasoning,84.5,https://storage.googleapis.com/deepmind-media/gemini/gemini_3-7_flash_model_evaluation.pdf,"{""context"":""not disclosed"",""dataset_split"":""1,000 validation questions"",""dataset_version"":""CharXiv Reasoning"",""effort"":""medium"",""harness"":""Google self-computed CharXiv evaluation"",""judge"":""not disclosed"",""mode"":""thinking"",""multimodal_input"":true,""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""pass@1"",""temperature"":""not disclosed"",""tools"":""none"",""trials"":""not disclosed""}",True,tech_report,verified,Information synthesis from complex charts. Physical rendering retained even when semantically identical to another official rendering.
gemini-3.7-flash,code_arena_frontend_elo,1588.0,https://storage.googleapis.com/deepmind-media/gemini/gemini_3-7_flash_model_evaluation.pdf,"{""aggregation"":""Arena Elo from pairwise human votes"",""context"":""not disclosed"",""dataset_split"":""live WebDev Arena"",""dataset_version"":""Code Arena WebDev live leaderboard"",""effort"":""medium"",""harness"":""Code Arena WebDev"",""judge"":""human pairwise preference votes"",""mode"":""thinking"",""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""not disclosed"",""temperature"":""not disclosed"",""tools"":""web application generation"",""trials"":""not disclosed""}",True,tech_report,verified,Elo. Physical rendering retained even when semantically identical to another official rendering.
gemini-3.7-flash,deep_swe_v1_1,65.3,https://storage.googleapis.com/deepmind-media/gemini/gemini_3-7_flash_model_evaluation.pdf,"{""context"":""not disclosed"",""dataset_split"":""113 tasks across 91 repositories"",""dataset_version"":""DeepSWE v1.1"",""effort"":""high"",""harness"":""Google self-computed mini-swe-agent with LiteLLM 1.96"",""judge"":""DeepSWE programmatic verifier"",""mode"":""thinking"",""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""pass@1"",""temperature"":""not disclosed"",""tools"":""mini-swe-agent repository shell/editor toolset"",""trials"":""not disclosed""}",True,tech_report,verified,Long-horizon software engineering. Physical rendering retained even when semantically identical to another official rendering.
gemini-3.7-flash,frontiercode_main_v1_1,43.6,https://storage.googleapis.com/deepmind-media/gemini/gemini_3-7_flash_model_evaluation.pdf,"{""context"":""not disclosed"",""dataset_split"":""100-task Main split"",""dataset_version"":""FrontierCode 1.1"",""effort"":""medium"",""harness"":""FrontierCode 1.1 official leaderboard"",""judge"":""blocking held-out functional criteria plus weighted maintainer-authored rubric criteria"",""mode"":""thinking"",""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""not disclosed"",""temperature"":""not disclosed"",""tools"":""coding agent with guarded internet access"",""trials"":""not disclosed""}",True,tech_report,verified,Score; production code quality. Physical rendering retained even when semantically identical to another official rendering.
gemini-3.7-flash,gdm_mrcr_v2_8needle_upto_128k,97.0,https://storage.googleapis.com/deepmind-media/gemini/gemini_3-7_flash_model_evaluation.pdf,"{""context"":""not disclosed"",""dataset_split"":""cumulative through 128K; 484 examples"",""dataset_version"":""GDM MRCR v2.1, 8 needles"",""effort"":""medium"",""harness"":""Google self-computed GDM eval_hub"",""judge"":""strict MRCR scorer"",""mode"":""thinking"",""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""pass@1"",""sequence_length"":""cumulative through 128K"",""temperature"":""not disclosed"",""tools"":""none"",""trials"":""not disclosed""}",True,tech_report,verified,Cumulative through 128K. Physical rendering retained even when semantically identical to another official rendering.
gemini-3.7-flash,gdp_pdf,34.0,https://storage.googleapis.com/deepmind-media/gemini/gemini_3-7_flash_model_evaluation.pdf,"{""aggregation"":""strict all_pass/mean"",""context"":""not disclosed"",""dataset_split"":""test"",""dataset_version"":""GDP.pdf 100-task held-out test set"",""effort"":""medium"",""harness"":""Google self-computed GDP.pdf evaluation"",""judge"":""Gemini 3.5 Flash rubric judge"",""mode"":""thinking"",""multimodal_input"":true,""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""pass@1"",""temperature"":""not disclosed"",""tools"":""none"",""trials"":""not disclosed""}",True,tech_report,verified,Strict task pass rate; expert PDF comprehension. Physical rendering retained even when semantically identical to another official rendering.
gemini-3.7-flash,gdpval_aa_v2_elo,1525.0,https://storage.googleapis.com/deepmind-media/gemini/gemini_3-7_flash_model_evaluation.pdf,"{""aggregation"":""blind pairwise Bradley-Terry Elo over one agentic submission per task"",""context"":""not disclosed"",""dataset_split"":""220 tasks"",""dataset_version"":""GDPval-AA v2"",""effort"":""medium"",""harness"":""Artificial Analysis Stirrup; fresh E2B sandbox; 250-turn limit"",""judge"":""panel of three frontier LLM judges sampled per comparison"",""mode"":""thinking"",""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""pass@1"",""temperature"":""not disclosed"",""tools"":""Web Fetch, Web Search, View Image, Code Exec, Finish, Abandon Task"",""trials"":1}",True,tech_report,verified,Elo; knowledge work. Physical rendering retained even when semantically identical to another official rendering.
gemini-3.7-flash,harvey_lab_aa,90.7,https://storage.googleapis.com/deepmind-media/gemini/gemini_3-7_flash_model_evaluation.pdf,"{""aggregation"":""rubric-criterion pass rate"",""context"":""not disclosed"",""dataset_split"":""120 tasks"",""dataset_version"":""Harvey LAB-AA"",""effort"":""medium"",""harness"":""Artificial Analysis Harvey LAB-AA"",""judge"":""rubric criteria over extracted deliverables"",""mode"":""thinking"",""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""not disclosed"",""temperature"":""not disclosed"",""tools"":""Stirrup shell and document workflow tools"",""trials"":""not disclosed""}",True,tech_report,verified,Complex legal workflows. Physical rendering retained even when semantically identical to another official rendering.
gemini-3.7-flash,hle_verified,53.6,https://storage.googleapis.com/deepmind-media/gemini/gemini_3-7_flash_model_evaluation.pdf,"{""context"":""not disclosed"",""dataset_split"":""Gold 668 plus Revision 1,143"",""dataset_version"":""HLE-Verified full verified set"",""effort"":""medium"",""harness"":""Google self-computed HLE-Verified evaluation"",""judge"":""HLE-Verified answer evaluator"",""mode"":""thinking"",""multimodal_input"":true,""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""pass@1"",""temperature"":""not disclosed"",""tools"":""none"",""trials"":""not disclosed""}",True,tech_report,verified,"1,811 items; Gold + Revision; Uncertain excluded. Physical rendering retained even when semantically identical to another official rendering."
gemini-3.7-flash,labbench2,82.1,https://storage.googleapis.com/deepmind-media/gemini/gemini_3-7_flash_model_evaluation.pdf,"{""context"":""not disclosed"",""dataset_split"":""full 1,912-task suite"",""dataset_version"":""LABBench2"",""effort"":""medium"",""harness"":""Google self-computed LABBench2 evaluation"",""judge"":""task-specific exact/programmatic evaluator"",""mode"":""thinking"",""multimodal_input"":true,""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""pass@1"",""temperature"":""not disclosed"",""tools"":""Linux terminal, bioinformatics tools, Python, R, and internet"",""trials"":""not disclosed""}",True,tech_report,verified,Biology real-world research tasks. Physical rendering retained even when semantically identical to another official rendering.
gemini-3.7-flash,lvbench,85.4,https://storage.googleapis.com/deepmind-media/gemini/gemini_3-7_flash_model_evaluation.pdf,"{""context"":""not disclosed"",""dataset_split"":""103 videos; 1,549 questions"",""dataset_version"":""LVBench"",""effort"":""medium"",""harness"":""Google self-computed LVBench evaluation"",""judge"":""exact multiple-choice answer"",""mode"":""thinking"",""multimodal_input"":true,""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""pass@1"",""temperature"":""not disclosed"",""tools"":""none"",""trials"":""not disclosed"",""video_frames"":1024}",True,tech_report,verified,Long video understanding. Physical rendering retained even when semantically identical to another official rendering.
gemini-3.7-flash,osworld_2_0_2026_06_24_partial,47.9,https://storage.googleapis.com/deepmind-media/gemini/gemini_3-7_flash_model_evaluation.pdf,"{""aggregation"":""maximum partial score over three single-attempt runs"",""context"":""not disclosed"",""dataset_split"":""108 tasks; before 2026-08-08 patch"",""dataset_version"":""osworld-v2-2026.06.24"",""effort"":""medium"",""harness"":""OSWorld 2.0 Docker with official Gemini/Anthropic CUA harness"",""judge"":""official weighted checkpoint partial scorer"",""max_steps"":500,""mode"":""thinking"",""multimodal_input"":true,""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""pass@1"",""screen"":""1080p screenshot-only observation"",""temperature"":""not disclosed"",""tools"":""pyautogui actuation plus UI-specific function declarations"",""trials"":3}",True,tech_report,verified,Weighted checkpoint partial score. Physical rendering retained even when semantically identical to another official rendering.
gemini-3.7-flash,terminal_bench_2_1,85.8,https://storage.googleapis.com/deepmind-media/gemini/gemini_3-7_flash_model_evaluation.pdf,"{""context"":""not disclosed"",""dataset_split"":""89 tasks"",""dataset_version"":""Terminal-Bench 2.1"",""effort"":""medium"",""harness"":""Terminus 2; Google self-computed"",""judge"":""Terminal-Bench 2.1 executable verifier"",""mode"":""thinking"",""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""pass@1"",""temperature"":""not disclosed"",""tools"":""Terminus 2 terminal toolset"",""trials"":""not disclosed""}",True,tech_report,verified,Agentic terminal coding. Physical rendering retained even when semantically identical to another official rendering.
gemini-3.7-flash,terminal_bench_3_0,14.9,https://storage.googleapis.com/deepmind-media/gemini/gemini_3-7_flash_model_evaluation.pdf,"{""context"":""not disclosed"",""dataset_split"":""74 tasks"",""dataset_version"":""Terminal-Bench 3.0 / FrontierBench"",""effort"":""medium"",""harness"":""Google self-computed mini-swe-agent with LiteLLM 1.96; two documented task compatibility modifications"",""judge"":""Terminal-Bench 3 task-success evaluator"",""mode"":""thinking"",""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""pass@1"",""temperature"":""not disclosed"",""tools"":""container terminal"",""trials"":""not disclosed""}",True,tech_report,verified,General agent capabilities. Physical rendering retained even when semantically identical to another official rendering.
gpt-5.6-terra,charxiv_reasoning,85.9,https://storage.googleapis.com/deepmind-media/gemini/gemini_3-7_flash_model_evaluation.pdf,"{""context"":""not disclosed"",""dataset_split"":""1,000 validation questions"",""dataset_version"":""CharXiv Reasoning"",""effort"":""not disclosed"",""harness"":""Google self-computed CharXiv evaluation"",""judge"":""not disclosed"",""mode"":""thinking"",""multimodal_input"":true,""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""not disclosed"",""temperature"":""not disclosed"",""tools"":""none"",""trials"":""not disclosed""}",True,tech_report,verified,Information synthesis from complex charts. Physical rendering retained even when semantically identical to another official rendering.
gpt-5.6-terra,code_arena_frontend_elo,1523.0,https://storage.googleapis.com/deepmind-media/gemini/gemini_3-7_flash_model_evaluation.pdf,"{""aggregation"":""Arena Elo from pairwise human votes"",""context"":""not disclosed"",""dataset_split"":""live WebDev Arena"",""dataset_version"":""Code Arena WebDev live leaderboard"",""effort"":""not disclosed"",""harness"":""Code Arena WebDev"",""judge"":""human pairwise preference votes"",""mode"":""thinking"",""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""not disclosed"",""temperature"":""not disclosed"",""tools"":""web application generation"",""trials"":""not disclosed""}",True,tech_report,verified,Elo. Physical rendering retained even when semantically identical to another official rendering.
gpt-5.6-terra,frontiercode_main_v1_1,41.3,https://storage.googleapis.com/deepmind-media/gemini/gemini_3-7_flash_model_evaluation.pdf,"{""context"":""not disclosed"",""dataset_split"":""100-task Main split"",""dataset_version"":""FrontierCode 1.1"",""effort"":""not disclosed"",""harness"":""FrontierCode 1.1 official leaderboard"",""judge"":""blocking held-out functional criteria plus weighted maintainer-authored rubric criteria"",""mode"":""thinking"",""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""not disclosed"",""temperature"":""not disclosed"",""tools"":""coding agent with guarded internet access"",""trials"":""not disclosed""}",True,tech_report,verified,Score; production code quality. Physical rendering retained even when semantically identical to another official rendering.
gpt-5.6-terra,gdm_mrcr_v2_8needle_upto_128k,93.5,https://storage.googleapis.com/deepmind-media/gemini/gemini_3-7_flash_model_evaluation.pdf,"{""context"":""not disclosed"",""dataset_split"":""cumulative through 128K; 484 examples"",""dataset_version"":""GDM MRCR v2.1, 8 needles"",""effort"":""not disclosed"",""harness"":""Google self-computed GDM eval_hub"",""judge"":""strict MRCR scorer"",""mode"":""thinking"",""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""not disclosed"",""sequence_length"":""cumulative through 128K"",""temperature"":""not disclosed"",""tools"":""none"",""trials"":""not disclosed""}",True,tech_report,verified,Cumulative through 128K. Physical rendering retained even when semantically identical to another official rendering.
gpt-5.6-terra,harvey_lab_aa,85.2,https://storage.googleapis.com/deepmind-media/gemini/gemini_3-7_flash_model_evaluation.pdf,"{""aggregation"":""rubric-criterion pass rate"",""context"":""not disclosed"",""dataset_split"":""120 tasks"",""dataset_version"":""Harvey LAB-AA"",""effort"":""not disclosed"",""harness"":""Artificial Analysis Harvey LAB-AA"",""judge"":""rubric criteria over extracted deliverables"",""mode"":""thinking"",""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""not disclosed"",""temperature"":""not disclosed"",""tools"":""Stirrup shell and document workflow tools"",""trials"":""not disclosed""}",True,tech_report,verified,Complex legal workflows. Physical rendering retained even when semantically identical to another official rendering.
gpt-5.6-terra,hle_verified,51.1,https://storage.googleapis.com/deepmind-media/gemini/gemini_3-7_flash_model_evaluation.pdf,"{""context"":""not disclosed"",""dataset_split"":""Gold 668 plus Revision 1,143"",""dataset_version"":""HLE-Verified full verified set"",""effort"":""not disclosed"",""harness"":""Google self-computed HLE-Verified evaluation"",""judge"":""HLE-Verified answer evaluator"",""mode"":""thinking"",""multimodal_input"":true,""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""not disclosed"",""temperature"":""not disclosed"",""tools"":""none"",""trials"":""not disclosed""}",True,tech_report,verified,"1,811 items; Gold + Revision; Uncertain excluded. Physical rendering retained even when semantically identical to another official rendering."
gpt-5.6-terra,labbench2,81.2,https://storage.googleapis.com/deepmind-media/gemini/gemini_3-7_flash_model_evaluation.pdf,"{""context"":""not disclosed"",""dataset_split"":""full 1,912-task suite"",""dataset_version"":""LABBench2"",""effort"":""not disclosed"",""harness"":""Google self-computed LABBench2 evaluation"",""judge"":""task-specific exact/programmatic evaluator"",""mode"":""thinking"",""multimodal_input"":true,""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""not disclosed"",""temperature"":""not disclosed"",""tools"":""Linux terminal, bioinformatics tools, Python, R, and internet"",""trials"":""not disclosed""}",True,tech_report,verified,Biology real-world research tasks. Physical rendering retained even when semantically identical to another official rendering.
gpt-5.6-terra,lvbench,78.9,https://storage.googleapis.com/deepmind-media/gemini/gemini_3-7_flash_model_evaluation.pdf,"{""context"":""not disclosed"",""dataset_split"":""103 videos; 1,549 questions"",""dataset_version"":""LVBench"",""effort"":""not disclosed"",""harness"":""Google self-computed LVBench evaluation"",""judge"":""exact multiple-choice answer"",""mode"":""thinking"",""multimodal_input"":true,""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""not disclosed"",""temperature"":""not disclosed"",""tools"":""none"",""trials"":""not disclosed"",""video_frames"":1024}",True,tech_report,verified,Long video understanding. Physical rendering retained even when semantically identical to another official rendering.
gpt-5.6-terra,osworld_2_0_2026_06_24_partial,50.2,https://openai.com/index/gpt-5-6/,"{""context"":""not disclosed"",""dataset_split"":""108 tasks; before 2026-08-08 patch"",""dataset_version"":""osworld-v2-2026.06.24"",""effort"":""not disclosed"",""harness"":""not disclosed"",""judge"":""not disclosed"",""mode"":""not disclosed"",""multimodal_input"":true,""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""not disclosed"",""temperature"":""not disclosed"",""tools"":""not disclosed"",""trials"":""not disclosed""}",True,official_blog,verified,Exact provider table captured from the rendered official page.
gpt-5.6-terra,terminal_bench_3_0,20.8,https://storage.googleapis.com/deepmind-media/gemini/gemini_3-7_flash_model_evaluation.pdf,"{""context"":""not disclosed"",""dataset_split"":""74 tasks"",""dataset_version"":""Terminal-Bench 3.0 / FrontierBench"",""effort"":""not disclosed"",""harness"":""historic official Terminal-Bench 3 leaderboard"",""judge"":""Terminal-Bench 3 task-success evaluator"",""mode"":""thinking"",""output_limit"":""not disclosed"",""prompt_style"":""not disclosed"",""sampling"":""not disclosed"",""temperature"":""not disclosed"",""tools"":""container terminal"",""trials"":""not disclosed""}",True,tech_report,verified,General agent capabilities. Physical rendering retained even when semantically identical to another official rendering.
claude-opus-5,mcpatlas_full_1000,85.8,https://research.meta.ai/articles/introducing-muse-code-and-muse-spark-1-2/evaluations/mcp-atlas-v1.png,"{""agent"":""Scale AI MCP-Atlas agent harness"",""attempts"":1,""configuration"":""500 public plus 500 held-out private tasks"",""context"":""not disclosed"",""effort"":""xhigh"",""harness"":""Scale AI MCP-Atlas agent harness and scoring pipeline"",""internet"":""containerized real MCP servers under benchmark controls"",""judge"":""claim-level 1/0.5/0; pass at coverage >=0.75; Gemini 3.1 Pro Preview primary"",""mode"":""thinking"",""prompt_style"":""not disclosed"",""sampling"":""pass@1 over all 1,000 tasks"",""temperature"":""not disclosed"",""tools"":""controlled target and distractor tools across 36 MCP servers and 220 tools""}",False,official_blog,verified,Exact printed value from the locked Meta release chart; evaluation settings come from the locked official methodology.
claude-opus-5,meta_internal_coding_bench,79.4,https://research.meta.ai/articles/introducing-muse-code-and-muse-spark-1-2/evaluations/meta-internal-coding-bench-v1.png,"{""agent"":""not disclosed"",""attempts"":2,""configuration"":""440 internal pull-request tasks"",""context"":""not disclosed"",""effort"":""max"",""harness"":""Meta internal agentic coding harness and isolated task sandbox"",""internet"":""disabled"",""judge"":""compile and unit tests in dedicated grading containers"",""mode"":""thinking"",""prompt_style"":""not disclosed"",""sampling"":""two attempts per task; average task-level success rate"",""temperature"":""not disclosed"",""tools"":""Meta internal agentic coding environment""}",False,official_blog,verified,Exact printed value from the locked Meta release chart; evaluation settings come from the locked official methodology.
claude-opus-5,terminal_bench_2_1,86.7,https://research.meta.ai/articles/introducing-muse-code-and-muse-spark-1-2/evaluations/terminal-bench-2-1-v1.png,"{""agent"":""Claude Code"",""attempts"":5,""configuration"":""selected official agent product in isolated Daytona"",""context"":""not disclosed"",""effort"":""max"",""harness"":""Meta internal agent evaluation framework; Claude Code; isolated Daytona sandbox"",""internet"":""not disclosed"",""judge"":""Terminal-Bench 2.1 executable task verifier"",""mode"":""thinking"",""prompt_style"":""not disclosed"",""sampling"":""pass@1 averaged across five attempts per task"",""temperature"":""not disclosed"",""tools"":""Claude Code terminal coding toolset""}",False,official_blog,verified,Exact printed value from the locked Meta release chart; evaluation settings come from the locked official methodology.
gemini-3.5-flash,mcpatlas_full_1000,83.6,https://research.meta.ai/articles/introducing-muse-code-and-muse-spark-1-2/evaluations/mcp-atlas-v1.png,"{""agent"":""Scale AI MCP-Atlas agent harness"",""attempts"":1,""configuration"":""500 public plus 500 held-out private tasks"",""context"":""not disclosed"",""effort"":""high"",""harness"":""Scale AI MCP-Atlas agent harness and scoring pipeline"",""internet"":""containerized real MCP servers under benchmark controls"",""judge"":""claim-level 1/0.5/0; pass at coverage >=0.75; Gemini 3.1 Pro Preview primary"",""mode"":""thinking"",""prompt_style"":""not disclosed"",""sampling"":""pass@1 over all 1,000 tasks"",""temperature"":""not disclosed"",""tools"":""controlled target and distractor tools across 36 MCP servers and 220 tools""}",False,official_blog,verified,Exact printed value from the locked Meta release chart; evaluation settings come from the locked official methodology.
gemini-3.6-flash,meta_internal_coding_bench,63.9,https://research.meta.ai/articles/introducing-muse-code-and-muse-spark-1-2/evaluations/meta-internal-coding-bench-v1.png,"{""agent"":""not disclosed"",""attempts"":2,""configuration"":""440 internal pull-request tasks"",""context"":""not disclosed"",""effort"":""high"",""harness"":""Meta internal agentic coding harness and isolated task sandbox"",""internet"":""disabled"",""judge"":""compile and unit tests in dedicated grading containers"",""mode"":""thinking"",""prompt_style"":""not disclosed"",""sampling"":""two attempts per task; average task-level success rate"",""temperature"":""not disclosed"",""tools"":""Meta internal agentic coding environment""}",False,official_blog,verified,Exact printed value from the locked Meta release chart; evaluation settings come from the locked official methodology.
gpt-5.6-terra,meta_internal_coding_bench,65.4,https://research.meta.ai/articles/introducing-muse-code-and-muse-spark-1-2/evaluations/meta-internal-coding-bench-v1.png,"{""agent"":""not disclosed"",""attempts"":2,""configuration"":""440 internal pull-request tasks"",""context"":""not disclosed"",""effort"":""max"",""harness"":""Meta internal agentic coding harness and isolated task sandbox"",""internet"":""disabled"",""judge"":""compile and unit tests in dedicated grading containers"",""mode"":""thinking"",""prompt_style"":""not disclosed"",""sampling"":""two attempts per task; average task-level success rate"",""temperature"":""not disclosed"",""tools"":""Meta internal agentic coding environment""}",False,official_blog,verified,Exact printed value from the locked Meta release chart; evaluation settings come from the locked official methodology.
kimi-k3,mcpatlas_full_1000,82.3,https://research.meta.ai/articles/introducing-muse-code-and-muse-spark-1-2/evaluations/mcp-atlas-v1.png,"{""agent"":""Scale AI MCP-Atlas agent harness"",""attempts"":1,""configuration"":""500 public plus 500 held-out private tasks"",""context"":""not disclosed"",""effort"":""max"",""harness"":""Scale AI MCP-Atlas agent harness and scoring pipeline"",""internet"":""containerized real MCP servers under benchmark controls"",""judge"":""claim-level 1/0.5/0; pass at coverage >=0.75; Gemini 3.1 Pro Preview primary"",""mode"":""thinking"",""prompt_style"":""not disclosed"",""sampling"":""pass@1 over all 1,000 tasks"",""temperature"":""not disclosed"",""tools"":""controlled target and distractor tools across 36 MCP servers and 220 tools""}",False,official_blog,verified,Exact printed value from the locked Meta release chart; evaluation settings come from the locked official methodology.
muse-spark-1.2,deep_swe_v1_1,59.3,https://research.meta.ai/articles/introducing-muse-code-and-muse-spark-1-2/evaluations/deepswe-1-1-v1.png,"{""agent"":""Muse Code"",""attempts"":5,""configuration"":""five-language Harbor conversion with pristine verifier"",""context"":""not disclosed"",""effort"":""xhigh"",""harness"":""Meta internal agent evaluation framework; Muse Code; isolated Daytona sandbox; Harbor conversion"",""internet"":""disabled during rollout and grading; model endpoint only"",""judge"":""functional verifier plus regression checks in a pristine verifier container"",""mode"":""thinking"",""prompt_style"":""not disclosed"",""sampling"":""pass@1 averaged across five attempts per task"",""temperature"":""not disclosed"",""tools"":""Muse Code repository shell/editor toolset""}",False,official_blog,verified,Exact printed value from the locked Meta release chart; evaluation settings come from the locked official methodology.
muse-spark-1.2,gdpval_aa_v2_elo,1631.0,https://research.meta.ai/articles/introducing-muse-code-and-muse-spark-1-2/evaluations/gdpval-aa-v2-v1.png,"{""agent"":""Artificial Analysis Stirrup"",""attempts"":1,""configuration"":""fresh E2B sandbox per task; 250-turn limit"",""context"":""not disclosed"",""effort"":""xhigh"",""harness"":""Artificial Analysis Stirrup; fresh E2B sandbox; 250-turn limit"",""internet"":""web fetch and Brave-backed web search available"",""judge"":""blind pairwise panel of three frontier LLM judges; Bradley-Terry Elo"",""mode"":""thinking"",""prompt_style"":""not disclosed"",""sampling"":""one agentic submission per each of 220 tasks"",""temperature"":""not disclosed"",""tools"":""Web Fetch, Web Search, View Image, Code Exec, Finish, Abandon Task""}",False,official_blog,verified,Exact printed value from the locked Meta release chart; evaluation settings come from the locked official methodology.
muse-spark-1.2,mcpatlas_full_1000,90.3,https://research.meta.ai/articles/introducing-muse-code-and-muse-spark-1-2/evaluations/mcp-atlas-v1.png,"{""agent"":""Scale AI MCP-Atlas agent harness"",""attempts"":1,""configuration"":""500 public plus 500 held-out private tasks"",""context"":""not disclosed"",""effort"":""xhigh"",""harness"":""Scale AI MCP-Atlas agent harness and scoring pipeline"",""internet"":""containerized real MCP servers under benchmark controls"",""judge"":""claim-level 1/0.5/0; pass at coverage >=0.75; Gemini 3.1 Pro Preview primary"",""mode"":""thinking"",""prompt_style"":""not disclosed"",""sampling"":""pass@1 over all 1,000 tasks"",""temperature"":""not disclosed"",""tools"":""controlled target and distractor tools across 36 MCP servers and 220 tools""}",False,official_blog,verified,Exact printed value from the locked Meta release chart; evaluation settings come from the locked official methodology.
muse-spark-1.2,meta_internal_coding_bench,70.6,https://research.meta.ai/articles/introducing-muse-code-and-muse-spark-1-2/evaluations/meta-internal-coding-bench-v1.png,"{""agent"":""not disclosed"",""attempts"":2,""configuration"":""440 internal pull-request tasks"",""context"":""not disclosed"",""effort"":""xhigh"",""harness"":""Meta internal agentic coding harness and isolated task sandbox"",""internet"":""disabled"",""judge"":""compile and unit tests in dedicated grading containers"",""mode"":""thinking"",""prompt_style"":""not disclosed"",""sampling"":""two attempts per task; average task-level success rate"",""temperature"":""not disclosed"",""tools"":""Meta internal agentic coding environment""}",False,official_blog,verified,Exact printed value from the locked Meta release chart; evaluation settings come from the locked official methodology.
muse-spark-1.2,terminal_bench_2_1,82.9,https://research.meta.ai/articles/introducing-muse-code-and-muse-spark-1-2/evaluations/terminal-bench-2-1-v1.png,"{""agent"":""Muse Code"",""attempts"":5,""configuration"":""selected official agent product in isolated Daytona"",""context"":""not disclosed"",""effort"":""xhigh"",""harness"":""Meta internal agent evaluation framework; Muse Code; isolated Daytona sandbox"",""internet"":""not disclosed"",""judge"":""Terminal-Bench 2.1 executable task verifier"",""mode"":""thinking"",""prompt_style"":""not disclosed"",""sampling"":""pass@1 averaged across five attempts per task"",""temperature"":""not disclosed"",""tools"":""Muse Code terminal coding toolset""}",False,official_blog,verified,Exact printed value from the locked Meta release chart; evaluation settings come from the locked official methodology.