ElevenLabs, a voice AI company, released an updated version of its music-generation model, Music v2, which can transition between genres during a track. The company claims the model is designed to manage both vocal and compositional complexity. The new release follows the initial launch of the music generation model almost 10 months ago. According to ElevenLabs, the model can shift from opera to heavy metal and back, produce fast rap without losing coherence, and incorporate non-musical sound effects into a track.
Artists can now select a part of a song and recreate it using prompts without affecting other sections of the track. Additionally, instead of generating short clips, creators can build a song by sections, such as the intro, verse, and chorus, then combine them. The model also performs more reliably across languages, lyrics, vocals, and arrangements. In recent months, AI labs have been competing to release models that can generate professional-grade music.
Companies like Google, Stability AI, and Suno have introduced new models capable of creating longer and more complex tracks. At Google I/O, the company added the ability to create covers, edit songs by sections, and generate music videos using its Flow Music tool. ElevenLabs highlighted that the new model is based on licensed data and is cleared for commercial use, allowing users to freely use the tracks. Negotiating with labels is critical, as other AI music startups, such as Suno and Udio, encountered legal issues over copyright concerns.
The model is now available on ElevenLabs’ ElevenCreative tool for marketing and branding teams, along with its newly launched ElevenMusic platform for creating AI-generated songs, with ElevenAPI availability coming soon.
Source: techcrunch
