Kokoro-82M
v1.0: 54 preset voices, 8 languages (US/UK English, Spanish, French, Hindi, Italian, Japanese, Brazilian Portuguese, Mandarin), 24 kHz. No voice cloning. v0.19 was 2024-12-25. ~11.5M HF downloads/month.
- Input
- text
- Output
- audio
- License
- apache-2.0
- Pricing
- per 1m characters: $0.65 (USD per 1M characters (cheapest hosted price per Artificial Analysis); self-hosting free) source
- Verified
- 2026-09-29
How to call it
| Provider | Model id | Endpoint / URL | Docs |
|---|---|---|---|
| Hugging Face | hexgrad/Kokoro-82M | huggingface.co/hexgrad/Kokoro-82M | — |
| pip | kokoro | github.com/hexgrad/kokoro | — |
| DeepInfra | hexgrad/Kokoro-82M | deepinfra.com/hexgrad/Kokoro-82M | — |
| OpenRouter | — | openrouter.ai/hexgrad/kokoro-82m | — |
Notable capabilities (1)
- Tiny model, top-tier quality: 82M-param StyleTTS2 + ISTFTNet model trained for ~$1,000 (1,000 A100 h) on permissive data; v0.19 hit #1 on the HF TTS Spaces Arena; still top-5 open weights on Artificial Analysis (~1065 Elo) in Sept 2026. source