F5 TTS

F5 TTS: Clone a voice. Narrate anything.

Upload a few seconds of a voice, paste its transcript, then paste your script of any length. F5 TTS runs locally on your device. Nothing is uploaded, there are no credits and there are no limits.

Preparing engine…

Diagnostics

Not loaded yet.

Script0 words

How F5 TTS works

F5 TTS is a voice cloning text to speech tool. You give it a short sample of a voice and it reads any script back in that voice. This page runs F5 TTS entirely in your browser, so it is free, private and has no credits.

Download the engine once. The first visit downloads about 1.4 GB. It is saved in your browser, so next visits start instantly.
Add a reference clip. Upload 5 to 12 seconds of clear speech from the voice you want to copy.
Type the reference text. Enter exactly what is said in the clip so F5 TTS can match words to sounds.
Paste your script and generate. Any length works. Press Play to listen and Download to save a WAV file.

Under the hood, F5 TTS splits your script into natural sentences. For each one it combines your reference clip, the reference text and the new text, then generates speech step by step using your graphics card through WebGPU (it falls back to the CPU, which is very slow). The parts are then stitched together with the pause and crossfade you choose. Speed, paragraph pause and crossfade sliders let you shape the result.

For best results use a clean clip with one speaker and no background noise. Generation time depends on your hardware.

Looking for simpler free options?

Try Free Text to Speech for ready-made voices in the cloud, or Free Text to Speech (GPU Based) to run ready-made voices on your own GPU.

Why use F5 TTS

F5 TTS voice cloning with a short clip

Five to twelve seconds of clean speech is enough to get started.

Free and private

F5 TTS runs on your device, so your voice clip and script stay with you.

Any script length

Long scripts are split into sentences and joined with the pause and crossfade you choose.

No sign-up

Download the engine once and start cloning, with no account and no credits.

What you can use F5 TTS for

Narrating a script in a voice you have permission to use
Creating a consistent voice for a series of videos
Voiceovers in your own voice
Prototyping character voices
Audiobook narration in a custom voice
Testing voice cloning without installing anything

Frequently asked questions

What is F5 TTS?

F5 TTS is a voice cloning text to speech model. You give it a short clip of a voice and the text of that clip, and it reads new text in a similar voice.

How long should the reference clip be?

Between 5 and 12 seconds of clean speech from one speaker with no background noise. Shorter or longer clips may not work well.

Why must I type the reference text?

F5 TTS matches the words in your clip to the sounds. The text should match exactly what is said.

Is anything uploaded?

No. F5 TTS runs on your own device. The engine files (about 1.4 GB) are downloaded once and saved in your browser.

Do I need a GPU for F5 TTS?

A modern GPU in Chrome or Edge with WebGPU is strongly recommended. The CPU fallback works but is very slow.

Can I clone anyone's voice?

Only clone voices you own or have clear permission to use. Do not use cloned voices to impersonate or mislead people.