This extension brings AnimateDiff into the AUTOMATIC1111 Stable Diffusion WebUI, turning the most popular image generation interface into an AI video toolkit. The open-source project has about 3,425 stars on GitHub and was most recently active in 2024. Once enabled, you can generate animated GIFs and short videos in almost exactly the same way you generate still images.
AnimateDiff works by inserting motion modules into the diffusion UNet. This extension implements that idea at runtime: the motion module is injected when you generate, so you do not need to reload or swap your model weights to switch between stills and animation. It integrates with ControlNet for guided motion and structure, supports prompt travel inherited from the animatediff-cli project so prompts can evolve across frames, and offers settings for context length, frame counts, and video output formats. Users of the Forge WebUI variant are directed to a dedicated branch that stays in sync. Because it lives inside the familiar WebUI, existing checkpoints, LoRAs, and workflows carry over to video work with very little relearning. It also fits naturally into existing WebUI habits: its settings live alongside the standard generation parameters, so animation experiments stay reproducible from the same seeds and settings used for still images.
The extension is free to use for non-commercial purposes under the license stated in the repository; commercial use requires contacting the author, as noted in the project README.
This extension is for Stable Diffusion users who want to move into AI video without leaving WebUI, and for creators producing short animated clips, loops, and GIFs from their existing models.