ElevenLabs lets artists fine-tune Music v2 with Vocals and Styles
ElevenLabs' new Vocals and Styles features fine-tune its Music v2 model on an artist's own tracks, generating songs in their voice and sound with copyright screening.
ElevenLabs on July 22 launched Vocals and Styles, a pair of personalization features that let musicians fine-tune the company's Music v2 model on their own recordings. Vocals is live in the ElevenMusic app, and Styles is available within Finetunes on ElevenCreative, the company's licensing-oriented platform.
Your voice, your sound, on demand
With Vocals, an artist uploads a selection of existing tracks and Music v2 fine-tunes to their vocal tone, texture, and delivery; every song generated afterward carries that voice. A faster one-shot path accepts a single voice recording and produces an original track sung in the uploader's voice, and users can alternatively pick from a pre-built Vocal Library. Styles applies the same idea to a whole sound: fine-tune on your catalog, or choose from a style library, and generate full tracks with consistent instrumentation and production character. Styles is now compatible with Music v2, bringing along that model's gains in vocal realism, arrangement, and multilingual delivery. ElevenLabs says all uploads are screened for copyright compliance before a fine-tune is created.
The release builds on Music v2, which shipped July 9 with long-form composition, mid-track genre transitions, and fast-rap capability, and extends ElevenLabs' push to make music a first-class product line alongside its voice business.
Why it matters
AI music has been dominated by prompt-to-song services whose training data provoked lawsuits from major labels, and rights-holders have converged on one demand: artist consent and control. ElevenLabs is building the consent story into the product — the artist supplies the training data, owns the resulting voice model, and generates with it deliberately, with copyright screening as a gate. If the approach holds up, fine-tuned personal models could become the standard commercial format for AI music: less a slot machine that emits generic songs, more an instrument an artist licenses and plays. That framing may prove as important commercially as any audio-quality gain, particularly as Suno and Udio negotiate their own settlements with the industry.