Voice cloning · local AI

Clone your own voice

Record ~20 seconds of speech and the local XTTS v2 model learns your voice. After saving, your voice appears in the editor as a base voice you can make say anything.

Record about 20 seconds in a quiet room. Your recording is used only to create your voice and stays private to your account — only you can generate speech with it.

Record your voice

Read the text below out loud for about 20 seconds in a quiet room — or upload an existing clip (.mp3 / .wav, up to 25 MB). The model analyzes a short sample to clone your voice.

Hello! I am recording a short sample of my voice so it can be cloned. I enjoy reading out loud, telling stories, and trying new technology. The quick brown fox jumps over the lazy dog, and the early bird catches the worm. This should be more than enough for a clear, natural sounding voice.
0s / 30s
or

My voices (0)

No cloned voices yet. Record one above to get started.