AI Voice Clone
Clone a voice from a short recording, then say anything in it.
Audio tool · MiniMax, Chatterbox and Lux TTS
Give it at least 10 seconds of a voice and a script. It reads the script back in that voice.
Listen to examples
“Hi, this is a quick reminder that your appointment is tomorrow at ten. Reply to this message if you need to move it.”
Reference audio
Chatterbox
Lux TTS
Voice Clone makes speech in a voice taken from a short sample. Upload a recording or record one in the browser (at least 10 seconds), type the text you want spoken, up to 1,000 characters, and choose one of three voice cloning models: MiniMax Voice Clone, Chatterbox or Lux TTS. The model listens to the sample for the tone, accent and character of the voice, then reads your script in it. Nothing is stored as a reusable voice profile; each generation works from the reference you attach, so you can switch samples freely. Results are AI generated and depend heavily on the sample: one speaker, clear speech and no music give the closest match. Clone only your own voice or a voice you have clear permission to use. The app does not verify who is in the sample, so that responsibility is yours.
What it does
Three voice cloning models
MiniMax Voice Clone aims for high fidelity and natural speech patterns, and runs noise reduction on your sample. Chatterbox leans expressive, with more range and personality. Lux TTS focuses on natural prosody. Try the same script on two and compare.
Upload or record a sample
Drop an audio file on Add Voice Recording, or press Record Voice and read for a few sentences. The page shows the length and warns if the sample is under the 10 second minimum before you can generate.
Up to 1,000 characters a run
Type or paste the script, with a live counter as you go. That is enough for a short ad read, an intro, a voicemail greeting or a few lines of dialogue per generation.
How it works
- 1
Add the reference audio
Upload or record at least 10 seconds of one person speaking. Fifteen to thirty seconds of clean, natural speech in a quiet room gives the model more to learn from than a short or noisy clip.
- 2
Write the script
Type the words you want spoken in the cloned voice, up to 1,000 characters. Punctuation shapes the pauses, so write it the way it should sound.
- 3
Pick a model and generate
Choose MiniMax Voice Clone, Chatterbox or Lux TTS, sign in if asked, and press Generate; the price shows on the button. The audio appears in your Library under Audio, ready to play and download.
Who it's for
For creators voicing their own content
Clone your own voice once you have a clean sample, then make intros, corrections and extra lines without setting up the mic again. Fix a fluffed sentence by typing it.
For small businesses
Turn the owner's or a team member's voice, with their go-ahead, into phone greetings, on-hold messages and short promo reads that sound like the business rather than a stock voice.
For video and dubbing workflows
Make a new line in a speaker's voice, with their permission, then sync it onto their footage with Lip Sync. Voiceover and Voice Changer cover the cases where a preset voice is fine.
More to try
Frequently asked questions
How long does the voice sample need to be?
At least 10 seconds; the page will not generate with a shorter clip. Longer samples of clear, natural speech from a single speaker usually give a closer match, so 15 to 30 seconds is a good target.
How much does a voice clone cost?
It depends on the model and the script length. Chatterbox and Lux TTS cost 70 credits for each started 300 characters, so 70 to 280 credits for up to 1,000 characters. MiniMax Voice Clone costs 480 credits plus 48 for each started 1,000 characters, so 528 for a full script. The Generate button shows the exact price.
Which voice cloning model should I choose?
Try Chatterbox or Lux TTS first; they cost less, so it is cheap to compare both on the same script. Use MiniMax Voice Clone when you want its high fidelity speech patterns and are fine with the higher price.
Is it legal to clone a voice?
Cloning your own voice, or a voice whose owner has given you permission, is the safe case. Using someone's voice without consent, or to deceive people, can break laws on privacy, fraud and publicity rights. The app does not check consent for you, so only upload voices you have the right to use.
Does the cloned voice get saved?
No voice profile is saved. Each generation uses the reference audio you attach, and the finished speech is kept in your Library. To make another line in the same voice, attach the same sample again.
Can I clone a voice from a song or a video?
Music, sound effects and more than one speaker confuse the model, so a clip pulled from a song or a busy video usually clones poorly. Use a clean recording of the person talking on their own, and only one you have permission to use.
Try it on your own audio
The tool is at the top of this page. The price shows on the Generate button before you start.
Back to the tool