Halfway through an edit, an ambient bed or transition sound is missing and the library has nothing suitable. Stable Audio generates audio in the browser: describe a scene and style to get effects or instrumental passages, and existing audio can be extended in length. An account is required, and an API is offered for developers. Before using output as an asset, check its license terms; published video also needs a synthetic-content label as required.
What it does
- Text descriptions generate sound effects and instrumental clips
- Extend or continue from generated audio
- Several versions per description for comparing tone and texture
- API access for your own asset pipeline
- Output drops into an editing project as background music
Who it is for
- A video needs ambience or a transition sound missing from the library
- Interactive content and games need batches of candidate tracks
- Describing sound in words instead of searching keywords
- Wiring audio generation into your own tool
How to get started
- Describe scene, material and mood, not just a style name
- Generate a short version to check tone before extending
- Listen for loudness and noise floor, and process again if needed
