FrontierThe story, in brief
Stability AI releases a new audio model that can create six-minute songs
Stability AI's Audio 3.0 now generates full songs on-device. Six minutes. No cloud.
The KeyNews take
Why it matters
Stability AI is pushing audio generation toward on-device inference, signaling a shift in how generative audio models will be deployed and reducing latency/cost barriers for developers building audio products.
The key facts
5 to knowStability Audio 3.0 released
Small model variant runs on-device
Generates two-minute tracks (headline claims six-minute capability)
On-device inference capability — reduces cloud dependency
May 2026 release date
Go to the source
TechCrunch AItechcrunch.com
Publisher excerpt: Stability Audio 3.0 small model can run on-device and generate two-minute long tracks