Stability AI tests Stable Video Diffusion for short videos without audio

Stability AI has announced Stable Video Diffusion, a tool that can create short videos without audio using generative AI. The tool is based on Stable Diffusion, the company’s image generator.

The company plans to make the tool available via the browser and a text-to-video interface. Stability AI says that the tool can be used for advertising, education and entertainment, among other things. Stable Video Diffusion can create 14 to 25 frames, which can be played back at frame rates of 3 to 30 fps. At the moment the tool is only able to create gifs of a few seconds.

The tool is currently only intended for research purposes and will only later be suitable for commercial purposes. It is not yet known when the tool will become available. The model behind Stable Video Diffusion also allows users to create 3D videos from a single image. The code behind Stable Video Diffusion is available on GitHub.