whatmodel.sh / tts / Pocket TTS (English)
Pocket TTS (English)
↗ share
+ contribute
Pocket TTS (English) is an open-weights text-to-speech model from Kyutai, released January 2026. Its published size is 0.1B parameters. Licence: cc-by-4.0.
Catalogue identity notes
kyutai/pocket-tts default English model (100M parameters, CPU-oriented). The default English weights have been updated in place (english.yaml now equals english_2026-09; the launch weights were english_2026-01). The voice-cloning repo is gated; without access the CLI falls back to kyutai/pocket-tts-without-voice-cloning. Weights CC BY 4.0, code MIT.
$ how it compares
Scored text-to-speech models released closest to Pocket TTS (English), out of 148 with a score, then more from Kyutai.
| # | model | score | where | ||
|---|---|---|---|---|---|
| 25 |  Qwen3-TTS 12Hz 1.7B Base | 65±1 | local | ||
| 41 |  Inworld TTS-1.5 Max | 62±3 | api $10.00/M chars | ||
| 44 | Qwen3-TTS 12Hz 0.6B Base | 62±2 | local | ||
| 74 |  LEMAS-TTS (w/ prosody, multilingual_prosody) | 52±6 | local | ||
| 84 | LEMAS-TTS (w/o prosody, multilingual_grl) | 50±6 | local | ||
| more from Kyutai | |||||
| 93 | Kyutai TTS 1.6B en/fr | 47±5 | local | ||
see all text-to-speech models · find the ones that fit your machine
$ where to run it
Other ways to run it
- Run it as published · needs ~0.23 GB · from github.com/kyutai-labs
uvx pocket-tts generate --text "Hello world, this is a test."
Official weights and documentation: huggingface.co/kyutai ↗
$ improve this page
Use a short GitHub form or edit the data through a pull request. Every number needs a source; here's which ones we can use.
data updated · sources and licences · json