Less wiring. More creating.
Explore Qwen Image 2.1 with 20 prepared creative setups in one organized ComfyUI workflow. English guides, example prompts and color-coded sections give you starting points for creating, editing and experimenting with your own images.
What’s included?
Image generation & editing: Text-to-image, instruction-based editing and classic img2img with adjustable denoise.
Multi-reference composition: Combine up to 10 reference images for characters, clothing, environments and objects.
Outfits, scenes & styles: Clothing transfer, background replacement, object insertion/removal, reference-based style transfer, plus prompt presets for lighting, poses and camera angles.
Posters & typography: Prepared setups for poster layouts and replacing text within existing images.
Transparency & targeted edits: RGBA generation, subject extraction, transparent PNG editing, mask-guided changes, annotation-guided editing and outpainting.
Storyboards & detail: Character-reference setups for scene development and a generative 2K detail pass..
Select your models once
Choose your Qwen image model, encoder and VAE in the central loader block—they feed every section that needs them. Comfy Kitchen Attention and a shared Qwen reference cache are already connected. Keep the central block active and enable the section you want to use.
The 20 sections combine technical workflows with task-specific prompt presets—not 20 separate models or unique underlying features. They are editable starting points for your ideas, not guarantees of identical results.
Workflow only; model weights are not included. Kitchen requires compatible ComfyUI dependencies, and its speed benefit varies.
The gallery thumbnails are promotional artwork, not direct outputs from this workflow.