A UK-based AI voice platform founded by DJ Fresh and Nico Pellerin, focused on vocal replacement for music production with rights management as a core principle rather than an afterthought. It licenses voices directly from artists and rights holders with consent, splits revenue 50/50, and has been showcased by WIPO as an example of responsible AI voice technology. Producers upload a clean acapella (or a fully mixed track via Stem-Swap, which separates stems first), choose from a roster of licensed voices, and generate a replacement vocal that can be downloaded or used directly inside a DAW via VST plugin. It also offers Instant Voice Cloning from 10–20 seconds of speech for quick referencing (positioned as complementary to, not a replacement for, flagship models), a Voiceover Studio 2.0 for script-based TTS with per-word editing, pauses, and expressions, and custom RVC model export compatible with AutoTune’s Metamorph plugin. Pricing includes a free 60-second trial, subscriptions from £6.99/month, pay-as-you-go at £11.99 per 15 minutes, Stem-Swap at 15 credits per swap (1 credit = 1 second), and £7.99 for a downloadable custom model. Independent review Zekai scores it 7.5/10, praising accuracy and time savings while flagging the credit system as confusing and potentially costly at volume.
Pros & Cons:
Pros:
Rights-first model licenses voices directly from artists with consent and splits revenue 50/50, which is materially different from tools that scrape voices; WIPO has featured it as a responsible AI voice case study, giving it institutional validation that marketing cannot buy; VST plugin integration means producers can swap vocals inside their DAW without round-tripping through a browser; Stem-Swap handles fully mixed tracks when no acapella exists, removing a common blocker; custom RVC model export compatible with AutoTune Metamorph, plus official Antares collaboration, extends its use beyond the web platform; backing from artists like Diplo and Rob Swire adds professional credibility.
Cons: The credit system is the main friction point—Zekai’s review and user feedback both note credits burn faster than expected on large projects, and the per-second billing for downloads adds a layer of cost modelling; public voice model count is not disclosed, so users cannot gauge roster breadth before signing up; the company is relatively small with no disclosed team size, which is a sustainability consideration for production dependencies; VAT is charged on top of listed prices, which can surprise buyers outside the UK; compliance posture beyond the rights framework is not publicly documented, so enterprise buyers should verify data handling directly.
Who is it for?:Music producers and songwriters who want to prototype vocal ideas with different voices without booking session singers; artists creating harmonies, stacked vocals, or multilingual versions from a single performance; sound designers needing character voices for games or podcasts with clear rights; DAW users who want voice swapping inside their existing workflow via VST.