Comfy 发布 Comfy Agent，可在 ComfyUI 内构建并迭代工作流

Today we are excited to introduce Comfy Agent. It is now available to everyone in Comfy Cloud, and it will be in Comfy Desktop in a few weeks. Describe what you want, and Comfy Agent can build, run, and iterate on workflows right in front of you!

Try Agent on Cloud

Generation got easy. Craft didn’t.

There are many ways to generate an image or a video now. But making something as specific as what’s in your mind, is still hard.

Great work comes from iterations: trying models, adjusting composition, fixing details, comparing approaches, and doing it again until it feels right.

ComfyUI has always been the tool for that level of control, but too much creative time still goes into finding tools, wiring workflows, debugging errors, and communicating technical terms to other creatives.

But you shouldn’t have to. We strongly believe creators shouldn’t have to choose between ambition and complexity. Comfy Agent is our answer to that.

A teammate right inside the app

Comfy Agent is right inside ComfyUI. It knows the latest models, ComfyUI extensions, and workflows. It can plan and evaluate techniques, build workflows, fix errors, find assets, and handle technical terms, so that you can focus on the creative idea itself.

You can create and use public or private skills while you work. You can also have up to 5 chats in parallel, with full chat history.

Things you can ask it today:

• “Update this same generation to compare 3 more video models.”

• “Explain what this workflow does and where it’s slow.”

• “Re-run my product-shot workflow on these 20 images.”

Try Agent on Cloud

Two of you on one canvas

The agent works directly on your canvas while you edit the workflow too. We designed it this way so both of you won’t have to wait for or interrupt each other. Real-time collaboration is fun:

• Reference another workflow, drag an asset from the library, or mention a node

• It can understand assets visually

• You can watch the graph assemble. Move things and it will notice

Every control ComfyUI gives you is still there. Open any node and check every parameter.

What creators are making

Three creators worked with Comfy Agent ahead of launch.

Animation by 852話 | Andidea: the story of an original character YUI. Characters and imagined worlds in one creative workflow. The agent helped shape the look, explore scenes, and compare generation results. One artist, 100 hours, working with Comfy Agent.

VFX use case by Doug Hogan: Isolate, Mask, and Refine. A shot taken from image to compositing passes. Isolate the subject, generate masks, and work with depth and normals in one workflow.

Try Agent on Cloud

Get started

1. Log in at cloud.comfy.org

2. Open the “Agent” panel to claim free initial chats.

3. Ask it for something real: “Take this product photo and make 20 on-brand ad variants in 4 aspect ratios.”

Comfy Agent is free to try and uses the Comfy Credits you already have. Check pricing here.

What’s next?

Local Comfy will soon be fully agentic. Your GPU, your models, your custom nodes. The agent walks you through setup, picks models that fit your hardware, and customizes your environment.

We are also working on agent-built custom nodes. The agent will bring customized editing tools for you.

We are also adding new features, including more model options, better public skills and knowledge, and more external integrations.

Feedback

Comfy Agent is currently in beta. We are actively developing new features and fixing issues. Tell us what works and what doesn't and it is the best way to help us prioritize the right things to build.

• Feedback: https://links.comfy.org/agentfeedback

• Discord: join our server and find us in #comfy-agent channel

As always, enjoy creating!