Signing in with a social account is taken as agreement to the copyright, voice rights and service terms. Your mobile number is asked for after the first sign-in.
Terms & Consent (copyright · voice rights)
1. Voice rights and voice-training consent
Own or lawfully licensed voice only — the voice you upload is your own or one you hold lawful rights to use.
Consent to use — you agree to the uploaded voice being used for voice-model training and voice synthesis (singing generation).
No impersonation or deepfakes — you will not generate or use voice for impersonation, deepfakes, defamation, deception or any unlawful purpose.
Where responsibility lies — legal responsibility for uploading or using voice you hold no rights to rests with the member; the service gives no warranty on this.
Exercising your rights — members may ask how their voice model and data are used, and request deletion.
On violation — the voice and content in question are deleted and use is restricted.
2. Copyright and uploaded-audio consent
Rights in the original recording you upload are the member's responsibility, as is separating or using audio you hold no rights to.
For commercial use of separation results, clearing the underlying copyright is up to the member.
3. Publishing tracks (showcase)
Opt-in publishing — Tracks you create are private by default and appear in the showcase, on track pages and in search engines only when you choose to publish each one.
License to the service — For each published track, you grant the service a non-exclusive, royalty-free license to play and display it in the showcase, on track pages and on service introduction pages, to the extent needed. The license ends when you make the track private (copies already held outside the service, such as search-engine caches, are beyond the service's control).
Your responsibility — The prompts, lyrics and uploaded material of a published track must not infringe anyone else's rights; you are responsible for them.
Credit — For each track you can show your display name or stay anonymous (anonymous by default).
Restrictions — The service may restrict or unpublish tracks that use sounds whose license hasn't been verified, tracks that have received repeated reports, or tracks suspected of violating these terms or the law.
What is published — Published tracks are offered as streaming (MP3) only; lossless audio, per-track audio and MIDI are not published.
These terms are a draft, not yet legally reviewed. The time, version and device of your consent are recorded on your account, and a revision of the terms requires consent again.