ElevenLabs, Stability AI Launch Music v2 and Stable Audio 3.0 Models This Week

This week, ElevenLabs and Stability AI launched new AI music generation models. ElevenLabs released Music v2, which enables genre-switching within tracks, section-by-section composition, and selective regeneration of specific parts. Stability AI released Stable Audio 3.0, a four-model family with open weights for three variants, generating music up to six minutes and twenty seconds long, with improved melodic coherence over longer outputs.

Both companies emphasize licensed training data amid ongoing copyright litigation involving Suno and Udio. Suno, valued at $2.45 billion with roughly 100 million users, remains the market leader in AI music generation. ElevenLabs, which reached $500 million in annual recurring revenue in April 2026, cut pricing for Music v1 and v2 by up to 50% on its API and 40% on its self-serve platform.