🧬
Voice Clone
Clone a voice from a short sample, then generate new speech in it.
Free, No Account
Voice Clone uses zero-shot voice cloning — upload a short sample recording (at least a few seconds, clear audio) along with exactly what was said in it, and generate new speech in that same voice. There's no lengthy training step: the sample is validated once, then reused for every generation.
Click to choose a .wav file
How it works
- 1Upload a short, clear voice sample (a few seconds is enough) and type exactly what was said in it.
- 2Click Create Voice Profile — this validates the sample, not a long training run.
- 3Type any new text and generate speech in the cloned voice.
What people use it for
- ✓ Cloning your own voice for consistent narration across videos
- ✓ Character voices for content from a single reference line
- ✓ Prototyping how a script sounds in a specific voice
Frequently Asked Questions
How long does training take?+
There's no real training wait — it's zero-shot cloning, so the voice profile is ready in seconds, not minutes.
Why do I need to type what was said in the sample?+
The model conditions on the sample's transcript, not just its audio — an accurate transcript noticeably improves the clone's quality.
What makes a good sample?+
At least 2-3 seconds, clear speech, minimal background noise — very quiet or very short samples are rejected with a specific reason.