Stability AI, the innovative force behind Stable Diffusion, has unveiled a new generation of audio models known as Stability Audio 3.0. The company asserts that its most advanced model is capable of generating professional-grade music compositions exceeding six minutes in duration.
The Stability Audio 3.0 suite introduces four distinct models: small SFX (459 million parameters), small (459 million parameters), medium (1.4 billion parameters), and large (2.7 billion parameters). The two smaller models are optimized for on-device sound and music creation, supporting generations of up to two minutes.
Significantly, both the medium and large models are designed to produce comprehensive compositions lasting 6 minutes and 20 seconds. These models are engineered to consistently maintain musical structure and melodic integrity, representing a more than twofold increase in length compared to Stable Audio 2.0, which was released in 2024.
In a move to foster broader accessibility and innovation, Stability AI is making the small SFX, small, and medium models available with open weights, allowing for public use and modification. This marks a substantial advancement over previous open versions, such as Stable Audio Open, released in 2024, which offered music generation up to 47 seconds.
Access to the large model, however, is restricted to the company's API and self-hosting paid services. Furthermore, businesses with annual revenues exceeding $1 million are required to secure an enterprise license for its use.
The landscape of AI-driven music generation is increasingly competitive, with numerous entities, including Google and ElevenLabs, introducing their own models and tools. Yet, as demonstrated by the ongoing legal challenges involving Suno and Udio, the long-term viability of these services may hinge critically on robust data licensing agreements and strategic partnerships with music labels.
Addressing these industry-wide concerns, Stability AI previously forged agreements with Warner Music Group and Universal Music Group last year to collaborate on model development and music creation tools. The company has affirmed that its latest collection of audio models is built upon fully licensed data.
The AI startup is also actively developing a new array of products specifically tailored for professional musicians, though specific features remain undisclosed. To spearhead Stability AI's professional music initiatives, Ethan Kaplan, formerly the chief digital officer at Universal Audio and Fender, has joined the company.
This strategic hiring reflects a broader trend within the AI industry, where companies are bolstering their credibility by recruiting experienced music executives. Earlier this year, Suno appointed former Merlin CEO Jeremy Sirota as its chief commercial officer, while ElevenLabs brought in Derek Cournoyer from indie music publisher Kobalt to serve as a strategy lead for its music business.
The Editorial Staff at AIChief is a team of professional content writers with extensive experience in AI and marketing. Founded in 2025, AIChief has quickly grown into the largest free AI resource hub in the industry.