You do not need to be a trained singer to release a song in your own voice anymore. Voice conversion models can take a professionally sung vocal and re-perform it in your timbre — your tone, your texture — while keeping the melody, timing and words of the original performance.
On Autunes this is wired into two places, depending on where you start.
Route 1: Jam — from hum to a song in your voice
Record your idea on the Jam page, enable vocals, and tick Sing it in my voice. Your recording does double duty: the AI uses its melody and timing to compose the track, and its sound as the voice sample for the final vocals. One flow, one recording, and the song that comes back is both your melody and your voice.
- Jam seed in your voice: 10 + 8 = 18 credits
- Full ~3 minute song in your voice: 30 + 8 = 38 credits
Route 2: the editor's Voice tool — re-sing any track
Open any song in the Autunes editor and pick the Voice tool. Upload a 5-30 second voice sample — speaking works, singing works better — and the AI re-sings the track's vocals to match that voice while keeping everything else. You can save voices you use often, so re-voicing your next song is one click.
Both routes keep your data yours: voice samples are used for your conversions, not to train public models.