[Feature]: ComfyUI Intergration

[david6666666]

🚀 The feature, motivation and pitch

ComfyUI Intergration

Alternatives

• after @dougbtv DALL-E compatible image generation endpoint #292 merged, https://github.com/dougbtv/comfyui-vllm-omni/ provides a example of using vLLM-Omni's image capabilities (Qwen-Image, Qwen-Image-Edit) directly within ComfyUI workflows.

I believe integrating vLLM-Omni into ComfyUI would provide users with a better experience, especially considering how widely ComfyUI is used in diffusion models.

Additional context

cc @dougbtv

Before submitting a new issue...

• Make sure you already searched for relevant issues, and asked the chatbot living at the bottom right corner of the documentation page, which can answer lots of frequently asked questions.

[lishunyang12]

This feature looks interesting to me and i would like to collaborate with you for this issue:)

[fhfuih]

I am also interested in this feature, and can take a took recently ~

[david6666666]

After offline confirmation, it was decided that @fhfuih would implement this. Thank you everyone for your help.

[david6666666]

• diffusion: Add ComfyUI plugin support for SGLang Diffusion sgl-project/sglang#15271
• https://github.com/SXQBW/ComfyUI-Qwen-Omni