Stability AI released Stable Audio 3.0 in May 2026, a new family of four audio models capable of generating professional-grade music tracks of more than six minutes — more than double the length supported by its predecessor, Stable Audio 2.0, released in 2024.
The four models in the lineup are: small SFX (459M parameters), small (459M parameters), medium (1.4B parameters), and large (2.7B parameters). The two small models are designed for on-device use and can generate audio of up to two minutes. The medium and large models can produce full compositions of 6 minutes 20 seconds while maintaining musical structure and melodic tone.
Stability AI is releasing the small SFX, small, and medium models as open weights, meaning anyone can use and modify them. The large model is available only through the API and paid self-hosting services. Companies with more than $1 million in revenue will need an enterprise license to use it.
The models are built on fully licensed data. Stability AI previously signed deals with Warner Music Group and Universal Music Group to develop music creation tools. The company also noted that its earlier open release, Stable Audio Open, was limited to tracks of up to 47 seconds — making the new open-weight models a significant expansion in capability.
Stability AI said it is developing a suite of products aimed at professional musicians, though it did not provide details on specific features. Ethan Kaplan, former chief digital officer at Universal Audio and Fender, is joining the company to lead that effort.
The release comes as several companies, including Google and ElevenLabs, are active in the AI music generation space. Licensing and label partnerships have emerged as a key issue in the sector, as ongoing legal battles involving Suno and Udio illustrate.
Source: TechCrunch