Stability AI Launches Stable Audio 2.5 for Professional Use
The new model generates three-minute musical compositions in seconds and introduces an audio inpainting feature for extending existing recordings.
Enterprise-Grade Audio Generation
Stability AI on September 10, 2025, launched Stable Audio 2.5, an audio generation model designed for professional and enterprise use. The company said the new model can generate structured, multi-part musical compositions up to three minutes long. A key technical update is its inference speed; the model can produce a full track in less than two seconds on a high-end GPU.
The model was trained on a fully licensed dataset, making its output commercially safe for use in advertising, games, and other applications, according to Stability AI.
New Audio Inpainting Feature
A significant new feature …
Archive Access
This article is older than 24 hours. Create a free account to access our 7-day archive.