ElevenLabs releases Music v2 AI model that can switch genres mid-song

In short: ElevenLabs has launched Music v2, a music making AI model that can change genres within the same song and lets users redo a section without altering the rest.

What happened

ElevenLabs, best known for AI voices, released a new version of its music generation system called Music v2. A “model” here means the software that learns patterns from many examples and then creates new audio, a bit like a very advanced autocomplete for sound.

The company says Music v2 can switch styles in the middle of a track, for example moving from opera to heavy metal and back. It can also handle fast rap, and add non-musical sound effects.

One practical change is editing by section. Users can select part of a song and regenerate only that piece using written instructions, while keeping the rest unchanged. It is similar to rewriting one paragraph in a document without touching the other pages.

ElevenLabs also says users can build longer songs by creating parts like an intro, verse, and chorus, then stitching them together. The company added that the tool works more reliably across languages, lyrics, vocals, and arrangements.

ElevenLabs said Music v2 is built on licensed data and is cleared for commercial use, meaning users can use the tracks in paid work. This matters because other AI music companies, including Suno and Udio, have faced copyright-related lawsuits.

Music v2 is available through ElevenLabs’ ElevenCreative tool for marketing teams and its ElevenMusic app. The company said access through its ElevenAPI, a way for other software to connect to ElevenLabs (like a plug that lets apps talk to each other), is coming soon.

Why it matters

Tools like this lower the time and cost needed to draft music for ads, videos, and social posts. They also raise ongoing questions about how AI music is trained and who owns the rights, so licensing and clear usage rules will likely keep getting attention.

Source: TechCrunch AI