Audio · Stability AI

Stable Audio 3.0

Open-weight music model, six-minute tracks

[ Overview ]

Stability AI's four-model Stable Audio family (Small, Small SFX, Medium, Large) generates tracks up to six minutes, with three of the four sizes released as open weights.

[ Pros ]

Mostly open-weight, self-hostable
No per-track fees for local generation
Up to 6-minute compositions

[ Cons ]

Largest model is API-only
Enterprise license needed above $1M revenue

[ Tags ]

music generationopen-weightlatent diffusionSFX