Agents can help you write strong image prompts, and — when ComfyUI is configured and the agent is image-enabled — you can generate images locally.

Generate

  1. On the Images page (or in Chat's image tools), describe an idea and create a prompt.
  2. Edit the prompt if you like.
  3. Click Generate — the image is created by your local ComfyUI and saved to your workspace.

Nothing generates automatically. Image generation only happens when you explicitly click Generate.

← Back to all docs