Key Highlights:
- ElevenLabs new Music v2 model can reportedly switch genres mid-song without breaking vocals or composition.
- ElevenLabs now lets users regenerate only specific song sections instead of redoing full tracks.
- The model supports section-by-section music creation including intros, verses, choruses, and transitions.
- ElevenLabs says Music v2 performs better across multiple languages, vocal styles, and arrangements.
- The company claims the model was trained on licensed data with commercial usage already cleared.
If you ever tried generating music, voice clones, or more with simple text prompts, you probably know ElevenLabs. In case you are unaware, the platform is widely used for text-to-speech, voice cloning, automatic video dubbing, and creating interactive conversational AI agents. Well, it seems ElevenLabs is now jumping deeper into the AI music race again with its new music generation model, Music v2.
ElevenLabs wants AI music to feel more like an output straight out of real studio
One of the biggest highlights of the Music v2 model is that it can switch genres in the middle of a track without messing up your track. As ElevenLabs notes, the music can jump from opera to heavy metal, slide into fast rap sections, then shift back again while still keeping vocals and composition coherent. You can now create tracks section by section, including intros, verses, choruses, and transitions, before stitching everything together into a complete composition.
Notably, the Music v2 model allows creators to isolate specific parts of a song and regenerate only those sections using prompts without affecting the rest of the track. Needless to say, that sounds much closer to modern editing software than a regular AI tool.
ElevenLabs also says the new model performs better across multiple languages, vocal styles, lyrics, and arrangements. The timing of the launch makes perfect sense for ElevenLabs to debut something like this. Over the past few months, nearly every major AI company has suddenly started pushing harder into music generation. Google recently expanded its Flow Music tools with AI-generated covers, music video generation, and section-based editing features during Google I/O. Meanwhile, other companies like Stability AI, Suno, and Udio are also racing toward longer and more improved AI-generated songs.
Also read: Lyria 3 AI Music Tool : Features, Prompts and How to Use
The company is also trying to avoid the AI music lawsuit mess
Not to forget, ElevenLabs appears to be keeping it safe from legal drama, too. The company has also announced that the new Music v2 model is trained on licensed data and is greenlighted commercial usage. It is a wise move by the company considering lawsuits around AI-generated music have been piling up for a while across industry. As for the availability, Music v2 is currently available via ElevenLabs’ ElevenCreative platform and its newer ElevenMusic service, while API access reportedly is coming later down the line.
