SwarmUI

SwarmUI is a web user interface for AI image generation. It also supports AI video models and some AI audio models. The README names Krea 2, Stable Diffusion and Flux for images, MiniMax H3, Wan and LTX-2 for video, and ACE-Step for audio. Formerly StableSwarmUI.

SwarmUI uses ComfyUI as its backend. The installer can install ComfyUI for you, and the full ComfyUI node graph is available in the Comfy Workflow tab. The AUTOMATIC1111 Stable Diffusion WebUI can be used as a legacy backend. The Generate tab builds workflows from simple settings. Generations can be shared across several GPUs and machines.

It runs locally on Windows, Linux and Mac computers with Apple M-series chips. The website recommends an Nvidia RTX 30xx or newer GPU. It can also run on Google Colab, Runpod or Vast.ai. By default it runs entirely on your own computer and does not collect data.

SwarmUI is free and open source under the MIT license, funded by donations. The README lists version 0.9.8 Beta.

Made for

People who generate AI images and video on their own computer, from beginners to advanced users. The docs also speak to ComfyUI users, users of other local Stable Diffusion interfaces, users of online AI services, and teams running large AI services.

Use cases

• Generate images from text prompts with local models
• Generate videos with models such as Wan and LTX-2
• Edit custom ComfyUI node workflows
• Compare settings side by side with the Grid Generator
• Split generations across several GPUs or machines
• Control generation from other programs through its HTTP API

An alternative to

• ComfyUI

  Puts a simple Generate tab of settings on top of the ComfyUI backend, with the full node graph one tab away, and splits work across several GPUs or machines.

• Stable Diffusion web UI (AUTOMATIC1111)

  Runs image and video models such as Flux, Wan and LTX-2 through the ComfyUI backend, and splits work across several GPUs or machines.