Stable Audio Open Small
Stable Audio Open Small is an earlier compact text-to-audio model for short clips. It produces stereo output at 44.1 kHz and includes an eight-step inference example.
Model design and access
The card separates training data for the autoencoder and diffusion model. The autoencoder uses Freesound and Free Music Archive recordings; the diffusion model uses Freesound. The T5-based text encoder conditions generation. Optimized Arm files are deployment forms of this same release.Stable Audio Open Small official model documentation
Practical use
Practical guidance. The short output window fits interface sounds, Foley prototypes and loops that will be edited after generation. Measure delay on the actual target device, including text encoding and decoding. Longer Audio 3 Small models offer a different duration range, so choose based on the sound and device constraints rather than the Small label alone.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Access | Downloadable |
| Maximum duration | 11 seconds.Stable Audio Open Small official model documentation |
| License | Stability AI Community License.Stable Audio Open Small official model documentation |
| Deployment | Arm CPU guidance and prebuilt LiteRT/TFLite files.Stable Audio Open Small official model documentation |
Related models
- Stable Audio 3.0 Small SFX
Stability AI · Stable Audio
A compact text-to-audio model for sound effects with downloadable weights.
Launch date not confirmedDownloadable - Stable Audio Open 1.0
Stability AI · Stable Audio
Stable Audio Open 1.0 converts English text prompts into variable-length stereo audio.
Launch date not confirmedDownloadable