Introduction

Welcome to Comfy with ComfyUI!

I made this site to do more than hand out workflows: I want people to see what generative AI makes possible, and to learn how ComfyUI itself works and what it can do.

Templates will get a model running, but if you want to extend things the way you want from there, you really do need to know the basics.

I think it can also be a good place for experienced users to keep up with how ComfyUI is evolving.


Sections

Switch the section tabs in the top-left to browse articles by theme. Here is a quick overview of the main sections:

🧭 Begin With ComfyUI

Covers ComfyUI setup and basic operations.

The UI has been changing very often lately. I'd like to fix the pages, but it will probably change again, so I'll update them once things settle down a bit.

🖼️ Data / Image Utilities

Covers how to prepare the images and data you use for generation. It includes editing and converting images, and ways to handle many images at once.

🔼 AI Capabilities

Generative AI comes in many forms, and each has different strengths. This section is designed so you can start from "what you want to do" and find "what kind of AI you can use."

It also gives a very simple explanation of how image generation AI works.

🗺️ Basic Workflows

Explanations of basic workflows. They're not that different from the ComfyUI templates, but they're laid out to be more minimal and easier to follow.

☹️ FAQ / Troubleshooting

Sorts out questions whose answers everyone now seems to take for granted, like "Why is 512px so common?" and "Seeds 1234 and 1235 are completely different."

It also collects errors and problems you're likely to run into with ComfyUI, with the cause and fix for each.


Workflow Copy Button

This site provides workflow JSON for almost every workflow article.

Not many people know this, but if you copy the workflow JSON text and press Ctrl + V on the ComfyUI canvas, the workflow is pasted right in.

Test workflow:

stable_diffusion_1_5.json

Make good use of it!


How to Face Generative AI

Generative AI opens up possibilities that never existed before, but it can do far less than we imagine.

Making an image that references several characters while keeping them consistent? Still not achieved. Making a 1-minute video? Can't.

When AI can't do what you want, you might think there must be some solution. In most cases, though, AI just hasn't progressed far enough yet to do what you're hoping for.

Don't rush. Enjoy AI while it's still immature. That's something you can only do now.


Contact

For correction reports, article requests, and feedback, use the "Site-related" section on the contact page (or post directly to GitHub Issues).

On desktop, you can also send from the bottom-right assistant icon (it may be unavailable in some environments).

If you want to contact the author directly, please use the contact page as well: /en/contact/.


License

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

Have fun and use them freely |д゚)


Into the World of ComfyUI

Now, let's learn together so we can use ComfyUI in a Comfy way!