A finished song, sung in your own voice
Most AI music gives you a stranger singing your words. Recording a short clip once means the finished track is you — which is the difference between something you show people and something you keep to yourself.
How it works
Record one clip
A short sample of you talking or singing. Once only — it is reused for everything you make afterwards.Write the song
A description or your own lyrics. Any genre.Get it back in your voice
The finished track is sung in your voice, in tune, whether or not you can sing.Keep or share it
Export the audio, or send a link.
What actually makes it good
- Record somewhere quiet. Room noise is the single biggest quality factor.
- Speak normally rather than performing — a natural clip transfers better than an impression of a singer.
- You do not need to be able to sing. Pitch is handled; the clip is for timbre.
- One good clip beats several rushed ones, and it is reused across every song you make.
Questions
Do I need to be able to sing?
No. The clip captures what your voice sounds like, not your pitch. The performance is generated in tune.
How long does the clip need to be?
Short — a few seconds of clear speech is enough.
Can I use someone else's voice?
Only with their permission. Cloning someone's voice without consent is not allowed.
Is my voice clip private?
It is tied to your account and used to generate your songs. See the privacy policy for detail.
Make one now
Free to start. iPhone and Android.