Stability AI Launches Stable Audio 3.0 for Longer Tracks
Stability AI has officially unveiled Stable Audio 3.0, a new suite of generative models designed to produce professional-grade audio. The most advanced iteration in this release allows users to generate cohesive musical tracks lasting up to six minutes and 20 seconds, more than doubling the duration capacity of the previous 2.0 version launched earlier this year.
The new lineup includes four distinct models tailored for different use cases:
- Small SFX: 459M parameters, optimized for sound effects.
- Small: 459M parameters, capable of on-device music generation up to two minutes.
- Medium: 1.4B parameters, supporting full-length compositions.
- Large: 2.7B parameters, the flagship model for complex musical structure and tone.
Expanding Access and Commercial Licensing
Stability AI is adopting a tiered distribution strategy. The small SFX, small, and medium models are being released with open weights, allowing for broader experimentation. This marks a significant evolution from the 2024 Stable Audio Open release, which was limited to 47-second clips.
Access to the large model is restricted to API usage and paid self-hosting services. Furthermore, the company has implemented an enterprise licensing requirement for organizations generating over $1 million in annual revenue.
Focus on Licensed Data and Industry Talent
As legal scrutiny intensifies around generative music startups—highlighted by high-profile litigation involving Suno and Udio—Stability AI is emphasizing its proactive approach to copyright. The company asserts that its new models were trained exclusively on fully licensed data, bolstered by strategic partnerships inked last year with Warner Music Group and Universal Music Group.
To further bridge the gap between AI and the music industry, Stability AI has recruited Ethan Kaplan, formerly the chief digital officer at Universal Audio and Fender. Kaplan will lead the company’s efforts in developing professional-grade music tools, a move that mirrors a broader trend of AI firms hiring veteran music executives to navigate the evolving market, such as Suno’s appointment of Jeremy Sirota and ElevenLabs’ hiring of Derek Cournoyer.