whatmodel.sh / tts / KittenTTS Nano 0.8
KittenTTS Nano 0.8
↗ share
+ contribute
KittenTTS Nano 0.8 is an open-weights text-to-speech model from KittenML, released February 2026. Its published size is 0.014B parameters. Licence: apache-2.0.
Catalogue identity notes
Comparison-only. Paradee Table 3 row 'KittenTTS nano 0.8 (Bella)', 14.0M params, 56.8 MB fp32 file; KittenML/kitten-tts-nano-0.8-fp32 @7a1db64 (created 2026-02-19, Apache-2.0) ships kitten_tts_nano_v0_8.onnx of 56,767,095 bytes. Distinct from kitten-tts-mini-0-8 (80M) and the int8 nano repo. Needed so the two new Paradee benchmarks reach 3 catalogue models; the run itself already has Kokoro-82M and Piper lessac medium.
Score within the Vapi Humanness · Bradley–Terry rating, 2026-09-25 snapshot comparison set only.
$ how it compares
The text-to-speech models ranked either side of KittenTTS Nano 0.8, out of 148 with a score.
| # | model | score | where |
|---|---|---|---|
| 99 |  Dia2 2B | 45±6 | local |
| 100 |  Amazon Polly Neural | 45±4 | api $16.00/M chars |
| 101 | KittenTTS Nano 0.8 | 44±5 | comparison only |
| 102 | GPT-SoVITS v2 | 44±4 | local |
| 103 |  MeloTTS English | 44±5 | local |
see all text-to-speech models · find the ones that fit your machine
$ what it's good at
other tests2 not in a task · 90% of its score
| test | result | vs the others in its run | counts |
|---|---|---|---|
| Paradee small-TTS comparison · 200 WikiText-103 sentences · UTMOS (utmos22_strong) | 4.01 / 5 ✎ | █······█···█ beats 0/2 | 57% |
| Paradee small-TTS comparison · 200 WikiText-103 sentences · Whisper-base WER | 5.5% ✎ | ▄··········█ beats 2/2 | 34% |
$ where to run it
No verified API or download recorded yet · add one
Official weights and documentation: huggingface.co/KittenML ↗
$ how the score was worked out · Paradee: Distilling Kokoro-82… 90%
Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works
| source | results | share of the score |
|---|---|---|
 Paradee: Distilling Kokoro-82M into an 8M-Parameter Single-Voice TTS Model (arXiv:2610.06817v1), Table 3 small-model comparison | 2 | 90% |
| prior · pull toward the middle | 10% | |
| weighted average | = 44.4 |
$ 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