Zero-shot cloning
A few seconds of speech is enough — no training or long recording sessions.
Free voice tool · runs in your browser
Record a short sample of your own voice, then type anything and hear it spoken back in a voice that sounds like you. The AI model runs on your own device, so your voice never leaves your browser.
Loading Clone My Voice…
How it works
Read the on-screen consent sentence aloud for 6–20 seconds. Only live recordings are accepted.
The voice model downloads once (about 1.5 GB) and learns your voice from the sample.
Type any English text and press Speak. Download the result as WAV.
Features
A few seconds of speech is enough — no training or long recording sessions.
Powered by the open Chatterbox model running in your browser on WebGPU.
Only voices recorded live in the tab can be cloned, after reading a consent sentence.
Make your AI voice calmer or livelier.
FAQ
No. VerboNest only clones a voice recorded live in your browser after you read a consent sentence, and file uploads are disabled. Cloning other people's voices without permission can be illegal and is often used for scams.
No. The model is downloaded to your browser and runs on your device. Your recording and the generated audio stay on your computer.
The voice model is about 1.5 GB. It downloads once and is cached; later visits skip the download but still need some time to start the model.
Chrome, Edge and Brave on a desktop with WebGPU work best; other devices fall back to a much slower CPU mode. Only English speech is supported for now.
Usually recognisably similar in tone and accent, but not identical. A quiet room and clear, natural speech give the best result.
More free tools