Welcome to Comfy with ComfyUI

This site is not just for sharing workflows. It is built to help people understand what generative AI can do, and how to operate ComfyUI itself.

Templates can get you started quickly, but if you want to extend things your own way, you still need the fundamentals.

I also hope this can be a useful place for experienced users to track how ComfyUI is evolving.

Recommended Models, September 2026

New models appear almost every week, but few balance quality, versatility, and stability this well. For now, these are the safe choices.

Image generation

Krea 2 A flexible all-rounder that handles both illustration and photorealism.
Anima A light, easy anime model. A good next step from SDXL-based models.

Image editing

Qwen-Image-2.1 Generation and editing in one model, and it can even output transparent images.

Video generation

MiniMax H3 Generates video and audio together, with quality close to closed models.

Latest updates

2026.10.2 Notes Added a page on using the Runpod card 2026.9.30 Notes Added ComfyUI-TypedDecision page 2026.9.28 Notes Added Conditioning Zero Out page 2026.9.21 Basic Workflows Added Qwen-Image-2.1 page 2026.9.19 Data / Image Utilities Added Loops page

All updates

About the author

nomadoor

I have been using ComfyUI since its early days, so I know a little about how it is meant to be used.

I build workflows with readability in mind (readable nodes). I would be glad if that way of thinking spreads and helps more people enjoy generative AI more deeply.

For corrections, article requests, or feedback, use the contact form or reach me on X (@noma_door).

License

All text, workflows, and images on this site are CC0. Use them freely.