Stability AI has announced the release of Stability Audio 3.0, a new family of audio models that can generate professional-grade music compositions lasting over six minutes. This development marks a significant improvement over its predecessor, Stable Audio 2.0, which was limited to shorter tracks.
Model Specifications
The Stability Audio 3.0 lineup includes four distinct models: small SFX and small, both with 459 million parameters; medium at 1.4 billion parameters; and large at 2.7 billion parameters. The small models are designed for on-device sound generation, capable of producing music up to two minutes long. In contrast, the medium and large models can create full compositions of up to 6 minutes and 20 seconds while maintaining musical structure and melodic quality.
The previous version, released in 2024, allowed for music generation only up to 47 seconds, highlighting the enhancements made in this latest iteration. While the smaller models will be available with open weights for public use and modification, the large model will only be accessible through API and self-hosting paid services. Companies with revenues exceeding $1 million will need to obtain an enterprise license to use the large model.
Strategic Collaborations and Legal Considerations
Stability AI's progress is supported by partnerships with major music labels, including Warner Music Group and Universal Music Group. These collaborations are essential, as the company asserts that its new audio models are built on fully licensed data. However, ongoing legal disputes involving companies like Suno and Udio emphasize the need to navigate data licensing and partnerships for the sustainability of music generation services moving forward.
Future Prospects
In addition to the new audio models, Stability AI is reportedly developing a suite of products aimed at professional musicians. While specific details remain undisclosed, Ethan Kaplan, the former chief digital officer at Universal Audio and Fender, has joined the company to lead this initiative. This strategic hire reflects a growing trend in the AI sector, where companies are increasingly recruiting music industry veterans to enhance their credibility and offerings.
As the market for AI-generated music expands, Stability AI faces competition from companies like Google and ElevenLabs. The ability to create longer, more complex musical compositions may give Stability AI a notable edge, especially as industry partnerships and licensing agreements become crucial for success.
The advancements seen in Stability Audio 3.0 elevate the potential for AI in music creation and highlight the importance of navigating the legal and commercial complexities of the music industry. As Stability AI and its competitors move forward, the implications for artists and music creators will be significant, paving the way for new forms of collaboration and creativity.
