ElevenLabs has unveiled Music v2, an advanced music generation model capable of switching genres mid-track. The new release is designed to handle complex vocals and composition, allowing artists to pick parts of songs and re-create them using prompts without altering other sections.

The model can transform from opera to heavy metal seamlessly, deliver fast rap without losing coherence, and even add non-musical sound effects to tracks. With Music v2, users can build a song by creating sections such as the intro, verse, and chorus, which can then be stitched together into a complete track.

Notably, ElevenLabs has stated that the model performs more reliably across languages, lyrics, vocals, and arrangements. This development comes at a time when AI labs are racing to release models capable of generating professional-grade music. Google, Stability AI, and Suno have also launched new music-generation models with similar capabilities.

As other AI music startups like Suno and Udio faced legal issues related to copyright, ElevenLabs emphasizes that its model is built on licensed data and cleared for commercial use, allowing users to freely utilize the tracks. The company has made the new model available through its ElevenCreative tool for marketing and branding teams, as well as via its newly launched ElevenMusic platform.