Stability AI launches new audio models capable of generating six-minute songs

May 20, 2026

AI Summary

Stability AI has introduced Stability Audio 3.0, a new family of audio models that can create professional-grade music lasting over six minutes. The models include small, medium, and large versions, with the latter available only through paid services. This release marks a significant advancement from the previous version, which could only generate shorter compositions.

Stability AI has released a new family of audio models called Stability Audio 3.0, which includes four models: small SFX, small, medium, and large.
The small models can generate sound and music for up to two minutes, while the medium and large models can create compositions lasting 6 minutes and 20 seconds.
The new models are built on fully licensed data, following previous partnerships with Warner Music Group and Universal Music Group.
The small SFX, small, and medium models are available with open weights for public use, while the large model is accessible only through API and paid services.
Stability AI is also developing products aimed at professional musicians, led by Ethan Kaplan, a former executive at Universal Audio and Fender.
The audio generation market is competitive, with other companies like Google and ElevenLabs also developing similar technologies, highlighting the importance of data licensing and partnerships in the industry.

stability aiaudio modelmusic generationon-devicegenerative music

Stability AI launches new audio models capable of generating six-minute songs

Related Stories

Satirical Ads Mimic Tech Campaigns in Public Transit

Malaysia's Prime Minister Anwar Ibrahim to Introduce AI Avatar for Public Engagement

Character.AI launches AI-generated microdramas with interactive features for users

JioStar Utilizes AI to Enhance User Engagement in Streaming Services