ElevenLabs released Music v2 in May 2026, a new version of its music-generation model that can shift genres within a single track — moving, for example, from opera to heavy metal and back — without losing coherence.
The model introduces several new capabilities for creators. Artists can select a specific section of a song and regenerate it using text prompts without altering the rest of the track. They can also build songs section by section — intro, verse, chorus — and stitch them together, rather than being limited to short generated clips. The model can also incorporate non-musical sound effects into a track and is designed to handle complex vocal arrangements, including fast rap delivery.
ElevenLabs said Music v2 performs more reliably across languages, lyrics, vocals, and arrangements compared to its predecessor. The first version of the company’s music-generation model launched roughly 10 months earlier.
The release comes as several AI companies compete to offer professional-grade music generation tools. Google, Stability AI, and Suno have each released new music models capable of generating longer and more complex tracks. At Google I/O, Google demonstrated section-based editing, cover creation, and music video generation through its Flow Music tool.
ElevenLabs stated that Music v2 is built on licensed data and cleared for commercial use. The company highlighted this distinction given that AI music startups Suno and Udio have faced copyright-related court cases over their training data.
Music v2 is available through ElevenLabs’ ElevenCreative tool, aimed at marketing and branding teams, and on the newly launched ElevenMusic platform for AI-generated song creation. Access via ElevenAPI is described as coming soon.
Source: TechCrunch