Streaming + cloning
Try Kyutai Pocket TTS free — unlimited, in your browser
Kyutai's Pocket TTS is a 100-million-parameter streaming model: audio starts playing while the rest is still being generated, and it can clone a voice from a short reference recording — all on CPU, entirely in your browser.
One-time download, stored in your browser and reused offline. Manage or remove it any time on the Models page.
Specifications
| Developer | Kyutai |
|---|---|
| Parameters | 100 million |
| Download size | 189 MB (int8) |
| Output | 24.0 kHz mono |
| Voices | 28 |
| Languages | English |
| Voice cloning | Yes |
| Streaming | Yes — audio starts before generation finishes |
| License | CC-BY-4.0 (weights), MIT (code) |
| Speed | First audio in ~1s, streams ahead of playback on a modern CPU. |
| Runtime | WebAssembly (any CPU) |
Source: kyutai/pocket-tts
Strengths
- +True streaming — near-instant first audio
- +Voice cloning from ~10 seconds of reference audio
- +English speech with streaming playback
Limitations
- −189 MB download before first use
- −English only today; Kyutai lists more languages as future work
- −Cloning quality depends heavily on the reference recording
Voices (10)
Compare
Kyutai Pocket TTS — questions
Is Kyutai Pocket TTS free to use?
Yes. The weights are CC-BY-4.0 (weights), MIT (code) and they run on your device.
What hardware does Kyutai Pocket TTS need?
Any modern browser — it runs on CPU via WebAssembly, no GPU needed. The download is 189 MB, cached after first use.
Is my text private when using Kyutai Pocket TTS here?
Yes. The model executes inside your browser tab — the text you synthesize is never transmitted to a server.
Can I use Kyutai Pocket TTS output commercially?
Review the current model, voice, and source-dataset terms before commercial use. The catalog lists CC-BY-4.0 (weights), MIT (code) for this model.