Qwen-Edit 2509 ComfyUI workflow
This isn’t a new base checkpoint so much as a “batteries included” workflow pack for running Qwen’s image-edit model (Qwen/Qwen-Image-Edit-2509) inside ComfyUI. The model card focuses on practical, high-control editing tasks: guided redraw using line/depth maps, selective inpainting via masks, pose-conditioned edits, multi-image compositing, outpainting/expansion, and even virtual try-on style workflows.
What makes it useful is the specificity: it calls out the exact ComfyUI node packages you’ll need, plus a companion LoRA (lightx2v/Qwen-Image-Lightning) the author expects you to use. If you try it, treat it like an integration recipe: install the listed nodes, drop the provided model_patches into your ComfyUI models/ directory, then start with one of the “single control” examples (like masks or outpainting) before moving on to multi-image fusion.
Quick stats from the listing feed: pipeline: image-to-image · 42 likes.
Source listing: https://huggingface.co/models?sort=modified