Stability AI launches Stable Audio 3.0 with up to six-minute tracks and open weights

Stability AI launched Stable Audio 3.0, a four-model audio lineup that can generate tracks up to six minutes long.

Three open-weights versions are available on Hugging Face, expanding access to text-to-audio, sound effects, and music generation.

The release adds variable-length control and editing tools, including inpainting, for more flexible audio creation.

Stable Audio 3.0 is trained on licensed data and allows commercial use under its community license, with enterprise terms for larger firms.