Stable Video Diffusion generates an animation after it is conditioned from an uploaded image. Basically, this means the AI model uses what’s in a still image to animate a video.