Instructions to use Phil2Sat/Qwen-Image-Edit-Rapid-AIO-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Phil2Sat/Qwen-Image-Edit-Rapid-AIO-GGUF with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="Phil2Sat/Qwen-Image-Edit-Rapid-AIO-GGUF",
	filename="Qwen2.5-VL-7B-Instruct-abliterated/Qwen2.5-VL-7B-Instruct-abliterated.Q4_K_M.gguf",
)

llm.create_chat_completion(
	messages = "\"Astronaut riding a horse\""
)

Notebooks
Google Colab
Kaggle
Local Apps

llama.cpp

How to use Phil2Sat/Qwen-Image-Edit-Rapid-AIO-GGUF with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf Phil2Sat/Qwen-Image-Edit-Rapid-AIO-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf Phil2Sat/Qwen-Image-Edit-Rapid-AIO-GGUF:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf Phil2Sat/Qwen-Image-Edit-Rapid-AIO-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf Phil2Sat/Qwen-Image-Edit-Rapid-AIO-GGUF:Q4_K_M

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf Phil2Sat/Qwen-Image-Edit-Rapid-AIO-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf Phil2Sat/Qwen-Image-Edit-Rapid-AIO-GGUF:Q4_K_M

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf Phil2Sat/Qwen-Image-Edit-Rapid-AIO-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf Phil2Sat/Qwen-Image-Edit-Rapid-AIO-GGUF:Q4_K_M

Use Docker

docker model run hf.co/Phil2Sat/Qwen-Image-Edit-Rapid-AIO-GGUF:Q4_K_M

LM Studio
Jan
Ollama
How to use Phil2Sat/Qwen-Image-Edit-Rapid-AIO-GGUF with Ollama:
```
ollama run hf.co/Phil2Sat/Qwen-Image-Edit-Rapid-AIO-GGUF:Q4_K_M
```

Unsloth Studio new

How to use Phil2Sat/Qwen-Image-Edit-Rapid-AIO-GGUF with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for Phil2Sat/Qwen-Image-Edit-Rapid-AIO-GGUF to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for Phil2Sat/Qwen-Image-Edit-Rapid-AIO-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for Phil2Sat/Qwen-Image-Edit-Rapid-AIO-GGUF to start chatting

Pi new

How to use Phil2Sat/Qwen-Image-Edit-Rapid-AIO-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf Phil2Sat/Qwen-Image-Edit-Rapid-AIO-GGUF:Q4_K_M

Configure the model in Pi

# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "Phil2Sat/Qwen-Image-Edit-Rapid-AIO-GGUF:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use Phil2Sat/Qwen-Image-Edit-Rapid-AIO-GGUF with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf Phil2Sat/Qwen-Image-Edit-Rapid-AIO-GGUF:Q4_K_M

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default Phil2Sat/Qwen-Image-Edit-Rapid-AIO-GGUF:Q4_K_M

Run Hermes

hermes

Docker Model Runner
How to use Phil2Sat/Qwen-Image-Edit-Rapid-AIO-GGUF with Docker Model Runner:
```
docker model run hf.co/Phil2Sat/Qwen-Image-Edit-Rapid-AIO-GGUF:Q4_K_M
```

Lemonade

How to use Phil2Sat/Qwen-Image-Edit-Rapid-AIO-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull Phil2Sat/Qwen-Image-Edit-Rapid-AIO-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.Qwen-Image-Edit-Rapid-AIO-GGUF-Q4_K_M

List all available models

lemonade list

compare

by flychiken - opened Oct 17, 2025

Discussion

flychiken

Oct 17, 2025

original edition:
8G VRAM ,70G RAM,0.5s to load the changed prompt word,50s 480*832_5step

gguf:
8G VRAM ,32G RAM,50s to load the changed prompt word,60s 480*832_5step

and the gguf character consistency is not good.

Phil2Sat

Owner Oct 17, 2025

•

edited Oct 17, 2025

seems okay, gguf in any quant and smaller quant is always a little bit slower in inference but the load time!?

personally i can't even test the original model:
i7- 4790k
32GB ddr3 maxed out
16gb amd instinct Mi25
256gb sata ssd
3tb hdd

thats what i got, yesterday i tried to rebuild the whole package with qwen-image-edit and bf16 model for better quant but RAM says NONONONONONO....

what are your ComfyUI start parameters?
python main.py --listen 0.0.0.0 --force-non-blocking --disable-smart-memory --reserve-vram 0.05 thats what i usually use

but im on linux, and if some app or sometimes "IF" i use the desktop is filling my VRAM it uses GTT, so my full 16gb are free. (running ComfyUI headless over ssh)

maybe its the text model too large you could try the original whats inside the AIO qwen_2.5_vl_7b_fp8_scaled.safetensors and see how it does

this are my first load times from an 2012 HDD...

second run:

xiaofengling

Oct 19, 2025

•

edited Oct 19, 2025

on my computer

amd 3600x
32g ddr4 3000
4060ti 8G

model:Q5_K_M
vae:pig_vae
clip:qwen2.5-vl-7b-instruct-abliterated Q4_K_M
8 steps
target size: 832*1216

7.36s/it
prompt execution in 68s

character consistency is good

Thank you for quantization, this runs quite well

Phil2Sat

Owner Oct 25, 2025

original edition:
8G VRAM ,70G RAM,0.5s to load the changed prompt word,50s 480*832_5step

gguf:
8G VRAM ,32G RAM,50s to load the changed prompt word,60s 480*832_5step

and the gguf character consistency is not good.

maybe try a smaller quant for the text encoder like the qwen2.5-vl-7b-instruct-abliterated Q4_K_M. or the normal fp8 scaled safetensors

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment