01Voiceover
Calm narration
English
Reference voice
0:00 / 0:00
AltVoice clone
0:00 / 0:00
This is a quick sample of my voice for testing.
AI Voice CloningClone any voice
Clone any voice
with AI
Free AI voice cloning online. Create realistic voice copies in seconds — for content, voiceover, localization, and more. No install, no signup.
Fast
Clone in seconds
Lifelike
Natural sound and intonation
Private
Your data, protected
20 languages
One engine for all
Free AI voice cloning, in your browser
Why we built this
No paywalls. No installs. No training on your audio.
AltVoice is a free AI voice cloner that runs entirely in your browser. Built on the open-source MOSS-TTS-Nano model, it lets you clone any voice from a short sample and generate speech in 20 languages. Your audio goes in, your audio comes out, the file is deleted in 24 hours.
20Languages
0$To start
24hSample TTL
MITOpen weights
Hear it for yourself
Each pair below is a short reference voice and the same voice speaking a brand-new sentence — generated by AltVoice.
02Podcast
Bright voice
English
Reference voice
0:00 / 0:00
AltVoice clone
0:00 / 0:00
Hello, I’m recording this to demonstrate my voice.
03Audiobook
Deep tone
English
Reference voice
0:00 / 0:00
AltVoice clone
0:00 / 0:00
History is written in the small choices people make every day.
How it works
- 01
Upload a sample
A short, clean recording of 5–15 seconds in any common audio format. Quality matters more than length.
- 02
Type your text
Up to 500 characters in any of the 20 supported languages.
- 03
Get the audio
Within a few seconds you’ll hear the same voice reading your text.
Tips for the best results
Voice cloning is mostly about the sample. Five minutes of care here beats hours of fiddling with the text.
Record in a quiet room
Background noise, music, and reverb confuse the model. A closed room with soft furniture is ideal.
Speak naturally
Avoid reading like a robot. A natural cadence transfers far better than over-enunciated narration.
Aim for 8–12 seconds
Short enough to stay clean, long enough to capture pitch range. Avoid clipping the start or end.
Use the source language
Cross-language cloning works, but the cleanest results come from a sample in the same language as your text.
One speaker only
If two voices overlap in the sample, the model averages them. Pick a clip with just one speaker.
Tweak temperature
Lower (0.5–0.7) for flatter, predictable narration. Higher (0.9–1.1) for more expressive output.
Made for
Use cases
Video voiceover
Narrate your videos and presentations in your own voice — no studio.
Audiobooks
Prototype chapters or produce personal audiobooks at home.
Podcasts
Patch missed lines or translate episodes into other languages.
Accessibility
Preserve a voice for people who are losing the ability to speak.
Games & bots
Give characters and assistants distinct, memorable voices.
Education
Lectures and study materials read in your instructor’s voice.
20 languages
One engine — your voice in any of them.
- English
- Русский
- Español
- Français
- Deutsch
- Italiano
- Português
- Polski
- Nederlands
- Türkçe
- العربية
- हिन्दी
- 中文
- 日本語
- 한국어
- Tiếng Việt
- ไทย
- Bahasa Indonesia
- Українська
- Čeština
Frequently asked
Is it really free?
Yes. You can clone a few voices without registering. After signing up, you get a generous daily free tier — no credit card required.
How long should the voice sample be?
5 to 15 seconds of clean speech with no background music or noise works best. Shorter clips work but the result may be less faithful.
Do you keep my uploaded files?
Guest uploads are deleted within 24 hours. Files for registered users are kept for 30 days, or deleted on request from your dashboard.
Can I use someone else’s voice?
Only with their explicit consent. Creating deepfakes of public figures or anyone without permission is not allowed by our Terms of Service.
Which audio formats are supported?
WAV, MP3, and M4A. Files up to 20 MB. We re-encode everything to 24 kHz mono internally before synthesis.
Where is my data stored?
On our servers in Europe. You can delete your account and all associated data at any time from the dashboard.
Can I use the cloned voice commercially?
Yes. The audio you generate is yours to use, provided you have the rights to the source voice you cloned.
Do you train your model on my audio?
No. Your audio is used only to fulfil your generation request and is not used for model training.
How long does generation take?
Usually 5–15 seconds depending on the length of the text and current load. Progress is shown live.
What languages work best?
English and Russian are first-class. The other 18 languages — Spanish, French, German, Chinese, Japanese, and more — work well; results are best when the sample is in the same language as the target text.
How do I clone my voice with AI?
Record a 5–15 second sample of yourself speaking clearly, drop it into the cloner above, type the text you want spoken, and press Clone voice. The whole flow takes under a minute.
Can I get free voice cloning online with no install?
Yes — AltVoice runs entirely in the browser. Nothing to download, no plugins, no account required for the first few generations.
What's the best AI voice clone tool?
It depends on what you need. AltVoice is built for fast, free, browser-based cloning with privacy by default. Paid tools like ElevenLabs offer studio-grade output and longer-form generation; we focus on the free, frictionless path.
Create an account
to keep your history
Free forever. Email is the only thing we ask for.