Free AI voice cloning: clone your voice in 15 seconds

Record a short sample in the browser, type a line, and hear it in your own voice a few seconds later. No upload, no studio, no card.

Clone my voice Free to try · no card

What it does

DeepVoice is an AI voice cloning tool that works from a short sample of your own voice. Record fifteen seconds in the browser, type any line, and it comes back spoken in your voice a few seconds later. No studio, no plugins, nothing to install.

It is the same voice clone that powers our deepfake video maker: clone your voice once and every talking video you make of yourself uses it.

How to clone your voice

  1. Record 15 seconds. Press record, talk normally about anything, press stop. You hear it back before anything is cloned.
  2. Confirm it is yours. One tick. The recording is the consent record, kept with the clone.
  3. Type a line, hit Generate. Up to 15 seconds of speech on the free plan, 60 seconds on Creator and Pro. Download the audio or share the link.

What you can use a cloned voice for

  • Voice-overs for videos, shorts and explainers without re-recording every take.
  • Reading out scripts, drafts and messages in your own voice when you cannot record.
  • A consistent narrator across a series when your real voice changes from day to day.
  • Talking videos of yourself: pair the clone with a short clip and type what you want said.

Getting a better clone

Most of the quality comes from the fifteen seconds you record, not from the model. A quiet room beats a good microphone: no music, no traffic, no second person talking. Speak at your normal pace rather than reading carefully, because the clone copies rhythm as well as pitch, and a stiff sample gives a stiff voice. Cover a bit of range in the sample — a question, a longer sentence — so the clone has something to work from when your script does the same.

On the writing side, punctuation is the only direction the clone takes. A full stop lands, a comma breathes, a question mark lifts the end of the line, and an exclamation mark reads as one. If a line comes out flat, repunctuate it before re-recording anything.

Questions people ask

Is the voice cloning free?

Yes. The free plan gives you a cloned voice and 10 voice generations a month with no card. Creator and Pro raise the allowance, drop the watermark and let you upload a voice file instead of recording live.

How much audio do I need to clone my voice?

Fifteen seconds. You record it in the browser, listen back, and confirm it. Longer samples do not make a noticeably better clone; a quiet room and a normal speaking pace do.

Can I clone someone else's voice?

Only your own, or a voice whose owner has given you clear permission you can show us. Cloning a public figure, a colleague or a family member without that is against the terms and gets the account closed.

What languages does it work in?

The clone speaks whatever you type. English is strongest; other languages work but accent and rhythm are less reliable.

Does it sound like me or like a robot?

It sounds like you on a decent phone call: your pitch, pace and accent, without background noise. Emotion follows the punctuation, so an exclamation mark reads as one.

How long does a clip take?

Voice clips come back in a few seconds. Talking videos take about a minute.

Clone my voice

See also: Deepfake video maker · AI talking avatars · Pricing