Genre-hopping, rap, sound effects: ElevenLabs raises the bar for AI music generation

ElevenLabs has launched Music v2, a new version of its AI music-generation model that can switch between genres within a single track. Ivan Mehta reports for TechCrunch that the model is built to handle complex vocals and compositions, arriving nearly ten months after the company released its first music-generation model.

According to ElevenLabs, the model can transition from opera to heavy metal and back, deliver fast rap while maintaining coherence, and embed non-musical sound effects directly into a track.

Section-by-section control

Music v2 gives artists more precise control over their work. Key capabilities include:

• Regenerating a specific section of a song using text prompts without affecting other parts
• Building a full song section by section (intro, verse, chorus) and stitching the parts together
• Improved consistency across languages, lyrics, vocals, and arrangements

ElevenLabs is entering a competitive field. Google, Stability AI, and Suno have all recently released music-generation models capable of producing longer and more complex tracks. At its Google I/O conference, Google introduced cover creation, section-based editing, and music video generation through its Flow Music tool.

One area where ElevenLabs is positioning itself carefully is licensing. The company states that Music v2 is trained on licensed data and cleared for commercial use. This matters in a sector where legal risk is real: AI music startups Suno and Udio have both faced copyright lawsuits.

Music v2 is currently available through ElevenLabs’ ElevenCreative tool, aimed at marketing and branding teams, and through the newly launched ElevenMusic platform for AI-generated songs. Access via ElevenAPI is expected soon.