Free AI voice generator

How to make a free AI voice

1. Type or paste your text into the box. Up to 5,000 characters at a time.
2. Pick a voice and a speed. Heart is the most natural of the American voices, and Emma is a good British one.
3. Press the button. The first time, the page downloads the voice model once, about 115 MB, and shows you how far along it is. After that the voice is saved in your browser and starts sooner.
4. Listen, then press Download WAV to save the file. It is yours to use.

Why it is free

This is not a trial and there is no credit counter. The voice is an open model called Kokoro, with 82 million parameters, small enough to run on an ordinary laptop or phone. Your browser downloads it from us once and runs it on your own device, so your text never reaches a server and we pay nothing each time you use it. That is the whole reason it can stay free with no sign-up.

Which languages it speaks

English only, in two accents: 12 American voices (7 women, 5 men) and 8 British voices (4 women, 4 men). The Kokoro model has voices in other languages, but the in-browser version used here is built for English, so we only offer what we tested. For other languages, use the Hermoso AI voiceover tool.

What it is good for

• Drafts and timing. Hear a script before you film it, and check how long it runs.
• Explainers and tutorials. A clear, even voice reading a screen recording.
• Accessibility. An audio version of a post or an article.
• Rough voiceovers for ads. Try a hook out loud before you commit to a final take.

Limits to know

• It is an 82-million-parameter model, so it sounds natural for its size but not as expressive as the largest paid voices. You cannot set emotion, whisper or shout.
• No voice cloning and no non-English languages.
• The first load is about 115 MB. On a slow connection or an old phone it takes longer, and an old device may speak slower than real time.
• Numbers, acronyms and brand names are read as the model guesses them. Spell a word the way it should sound if it gets one wrong.

Need more than this?

For premium voices in many languages, cloned voices, music and sound, and a voiceover that goes straight onto a video ad, use Hermoso AI voiceover. The free voice and the premium voices sit side by side, so you can draft here and finish there. How the free tools compare with paid voice tools is in free AI voice generator vs paid.

Built on open models and code: licenses, credits and source.

Frequently asked questions

Is there a free AI voice generator with no sign-up?

Yes. This page runs an open text-to-speech model in your browser. You type text, press a button and download the audio as a WAV file, with no account and no limit on how many times you use it.

Is it really free?

Yes. The model runs on your own device, so it costs us nothing each time you use it and there is no credit counter, trial or watermark.

Is my text uploaded anywhere?

No. The text is spoken by a model running in your browser. Only the model files are downloaded, once, from assets.hermoso.ai.

Which languages does it support?

English only, with American and British accents: 12 American and 8 British voices. For other languages, use the Hermoso AI voiceover tool.

Why does it download about 115 MB the first time?

That is the voice model and the engine that runs it. They are saved in your browser, so the next time the voice starts without downloading again.

Can I use the audio commercially?

The Kokoro model is released under the Apache 2.0 license, which allows commercial use. Make sure the text you read aloud is yours to use.

What file do I get?

A WAV file, 24 kHz mono, 16-bit, that plays in any video or audio editor.

How is it different from premium AI voices?

It is a small open model, so it is less expressive and English only. Premium voices add many languages, emotion control and cloned voices, and they run on our servers rather than your device.