Stability AI releases a new audio model that can create six-minute songs

Stability AI, renowned for its work on Stable Diffusion, has announced the launch of a new series of audio models known as Stability Audio 3.0. This innovative lineup includes a flagship model capable of producing high-quality music tracks lasting over six minutes, marking a significant advancement in audio generation technology. The Stability Audio 3.0 family features four distinct models: small SFX (459 million parameters), small (459 million parameters), medium (1.4 billion parameters), and large (2.7 billion parameters). The smaller variants are designed for on-device sound and music generation, allowing for compositions of up to two minutes. In contrast, the medium and large models can deliver full musical pieces lasting an impressive 6 minutes and 20 seconds, surpassing the capabilities of the previous version, Stability Audio 2.0, which was launched in 2024. Stability AI is providing the small SFX, small, and medium models with open weights, enabling users to freely utilize and modify them. This marks a notable upgrade from the earlier versions, which allowed music generation of only up to 47 seconds. The larger model, however, is restricted to access via API and self-hosting paid services, with an enterprise license required for companies earning over $1 million annually. As the competition heats up, with companies like Google and ElevenLabs entering the music generation arena, the importance of licensing data and collaborating with music labels has become apparent, as evidenced by ongoing legal disputes involving Suno and Udio. Stability AI has proactively secured partnerships with major players like Warner Music Group and Universal Music Group to enhance their models and music creation tools. The company has emphasized that the new audio models are based on fully licensed data. Additionally, Stability AI is working on a suite of products aimed at professional musicians, although specific details have yet to be revealed. In a strategic move, Ethan Kaplan, the former chief digital officer at Universal Audio and Fender, will lead the development of Stability’s professional music offerings. Other AI firms are also bolstering their teams with music industry veterans; for instance, Suno recently appointed former Merlin CEO Jeremy Sirota as chief commercial officer, while ElevenLabs brought on Derek Cournoyer from Kobalt as a strategy lead for its music initiatives.