Stable Audio 2.5
Stability AI
Stability AI's enterprise audio model (September 2025) for high-quality sound and music production at scale. A generation breakthrough cuts inference from 50 steps to 8, producing three-minute tracks in seconds, and it supports text-to-audio, audio-to-audio, and audio inpainting on user files.
Strengths
- Enterprise-grade audio and music production at scale
- 8-step generation (down from 50) - tracks in seconds
- Up to three-minute compositions
- Text-to-audio, audio-to-audio, and audio inpainting
- Sub-2-second inference on a GPU
Caveats
- Proprietary - API and enterprise licensing
- Tuned for brand/enterprise sound over consumer songs
- No open weights
- Vendor
- Stability AI
- License
- Proprietary
- Release Date
- 2025-09-11
- Modalities
- textaudio
Capabilities
Vision
Audio
Video
Tool Use
Resources
Pricing
Usage-based via the Stability AI API and platform partners (fal); enterprise licensing for brands. Sub-2-second generation on a GPU.