Stability AI, the creator of Stable Diffusion, has unveiled a new series of audio models called Stability Audio 3.0. The leading model can produce professional-grade music exceeding six minutes, according to the company.
The release includes four models under Stable Audio 3.0: small SFX (459M parameters), small (459M parameters), medium (1.4B parameters), and large (2.7B parameters). The smaller models are designed for generating on-device sound and music up to two minutes in length.
The medium and large models can generate complete compositions lasting 6 minutes and 20 seconds, maintaining musical structure and melody. This duration is over twice what Stable Audio 2.0 could produce in 2024.
Stability AI is providing open weights for the small SFX, small, and medium models, making them available for anyone to use and modify. In 2024, the company introduced Stable Audio Open, allowing for music generation up to 47 seconds. The latest models represent a significant advancement from prior open versions.
The large model is accessible only via API and paid self-hosting services. Companies with annual revenue exceeding $1 million must acquire an enterprise license.
Various companies, such as Google and ElevenLabs, are developing music generation models and tools. However, ongoing legal challenges, as experienced by Suno and Udio, highlight that data licensing and partnerships with music labels may play a crucial role in the sustainability of these services.
Last year, Stability AI partnered with Warner Music Group and Universal Music Group to create models and music tools. The latest audio models are built on fully licensed data.
The AI startup is also working on new products for professional musicians, though details are scarce. Ethan Kaplan, former chief digital officer at Universal Audio and Fender, will join the company to lead its professional music offerings.
Other AI firms are similarly enhancing their credibility by hiring music executives. Earlier, Suno appointed former Merlin CEO Jeremy Sirota as chief commercial officer. ElevenLabs brought on Derek Cournoyer from indie publisher Kobalt as a strategy lead for its music business.