Artificial intelligence (AI) audio corporations ElevenLabs said on the 22nd it has released its next-generation AI music generation model, "Music v2."
Music v2 improves performance and quality in vocal precision, instrumentation, and arrangement over Music v1. It adds a feature that naturally transitions between different genres within a single track, such as from opera to heavy metal, and it renders fast rap stably. It can also blend sound effects (SFX) and background sounds naturally into the music, enabling the creation of more polished tracks.
Users can select and edit only specific sections of a track, or build the intro, verse, and chorus step by step to complete a long-form piece. Going forward, the company plans to add a feature that generates music in a similar style based on a reference clip of up to 30 seconds.
The newly unveiled Music v2 is offered through ElevenLabs' three core platforms: ElevenMusic, a workspace for general creators; Eleven API for developers; and ElevenCreative Music, a platform for advertisers and brand video producers.
To attract users, ElevenLabs cut fees on the Eleven API platform by up to 50% with the launch of Music v2. Fees for ElevenCreative will also be reduced by up to 40%.
Hong Sang-won, ElevenLabs' Korea head, said, "The new model is based strictly on fully licensed data and is a commercial AI music generation solution designed with copyright in mind, making it the safest and most powerful tool that can be freely used in the global market."