ComfyUI-Qwen-VL
A ComfyUI extension for Qwen-VL series large language models, supporting multi-modal functions such as text generation, image understanding, and video analysis.Support…
ComfyUI-Qwen-VL 🐼
<p align="center"> <a href="README_zh.md">中文</a>  ｜   English </p>Where Figma meets VSCode: Artistic vision meets engineering precision —— a romantic manifesto from designers to the code world.
✨ A ComfyUI extension for Qwen2.5-VL series large language models, supporting multimodal capabilities such as text generation, image understanding, and video analysis. ✨
🌟 Features
- Supports Qwen2-VL, Qwen2.5-VL and other series models
- Provides various functional nodes for text generation, image understanding, video analysis, etc.
- Supports model quantization configuration to optimize memory usage
- Offers an intuitive user interface for easy parameter adjustment
🚀 Installation
- Navigate to the
custom_nodesdirectory of ComfyUI - Clone this repository:
git clone https://github.com/SXQBW/ComfyUI-Qwen-VL.git - Install dependencies:
cd ComfyUI-Qwen-VL pip install -r requirements.txt - Restart ComfyUI
📖 Usage
- In the ComfyUI interface, locate the Qwen-VL related nodes
- Select the model and quantization method you want to use
- Configure generation parameters such as temperature and maximum tokens
- Connect input (text, image, or video) and output nodes
- Run the workflow
📦 Supported Models
The following models are currently supported:
- Qwen/Qwen2.5-VL-3B-Instruct
- Qwen/Qwen2.5-VL-3B-Instruct-AWQ
- Qwen/Qwen2.5-VL-7B-Instruct
- Qwen/Qwen2.5-VL-7B-Instruct-AWQ
- Qwen/Qwen2.5-VL-32B-Instruct
- Qwen/Qwen2.5-VL-32B-Instruct-AWQ
- Qwen/Qwen2.5-VL-72B-Instruct
- Qwen/Qwen2.5-VL-72B-Instruct-AWQ
- Qwen/Qwen2-VL-2B
- Qwen/Qwen2-VL-2B-Instruct
- Qwen/Qwen2-VL-7B-Instruct
- Qwen/Qwen2-VL-72B-Instruct
- Qwen/Qwen2-VL-2B-Instruct-AWQ
- Qwen/Qwen2-VL-2B-Instruct-GPTQ-Int4
- Qwen/Qwen2-VL-2B-Instruct-GPTQ-Int8
- Qwen/Qwen2-VL-7B-Instruct-AWQ
- Qwen/Qwen2-VL-7B-Instruct-GPTQ-Int4
- Qwen/Qwen2-VL-7B-Instruct-GPTQ-Int8
- Qwen/Qwen2-VL-72B-Instruct-AWQ
- Qwen/Qwen2-VL-72B-Instruct-GPTQ-Int4
- Qwen/Qwen2-VL-72B-Instruct-GPTQ-Int8
- huihui-ai/Qwen2.5-VL-7B-Instruct-abliterated
Quantization Options
- 🚫 None (Original Precision): Use original precision
- 👍 4-bit (VRAM-friendly): Use 4-bit quantization to save VRAM
- ⚖️ 8-bit (Balanced Precision): Use 8-bit quantization for balanced precision and performance
👀 Example Workflows
Here's a simple example workflow for image understanding:
📖 FAQ
Model Loading Issues
If you encounter errors loading the model, ensure:
- The model file path is correct
- You have sufficient GPU memory (Choose an appropriate model based on your VRAM size. Don't jump straight to the 72B model – brute force won't work here, it'll just crash your VRAM)
- All necessary dependencies are installed
About Quantization
When using pre-quantized models (e.g., AWQ versions), you may see the following warning: "Model Qwen2.5-VL-3B-Instruct-AWQ is already quantized, user quantization settings will be ignored." This is normal, and the plugin will automatically use the model's pre-quantized version.
🤝 Contributing
Contributions, issues, and feature requests are welcome!
🙏 Acknowledgments
Special thanks to the Qwen team for developing these powerful models, and to the ComfyUI community for their support!
The star you're about to click ✨
Is not just a gesture of approval, but a cosmic explosion where design thinking meets the code universe. When an artist's aesthetic obsession collides with a programmer's geek spirit – this might just be the most romantic chemical reaction on GitHub.