hexgrad · Voice model

Kokoro 82M

hexgrad/kokoro-82m

Voice

Released Apr 23, 2026

Kokoro 82M is a lightweight, open-weight text-to-speech model from hexgrad. It converts text to speech across 8 languages (American and British English, Spanish, French, Hindi, Italian, Japanese, Portuguese, and Chinese)...

Specifications

Input: text, up to 20,000 characters per call
Output: mp3 (default) or raw pcm
Voice: provider voice ids via `voice`; omitted = the provider's default when it has one
Billing: per character of input text

Use it

POST /api/v1/audio/speech · MCP speak · full reference

curl https://omnirail.org/api/v1/audio/speech \
  -H "Authorization: Bearer $OMNIRAIL_KEY" -H "Content-Type: application/json" \
  -d '{"model":"hexgrad/kokoro-82m","input":"Hello from OmniRail. Every model, one key.","response_format":"mp3"}' \
  --output hello.mp3