Overview
Stable Video Diffusion is Stability AI's generative video model based on the Stable Diffusion family. The official product page describes text-to-video generation, 14- and 25-frame outputs, adjustable frame rates, and deployment on a user's own infrastructure. Stability AI's knowledge base points users to the Generative Models repository for the reference implementation and to separately hosted model weights. The reference code is available under the MIT license, while weights and commercial use follow Stability AI's applicable terms. The hosted SVD API was deprecated in July 2025.