Ponkotsu Voice
● LiveA lightweight text-to-speech service that turns your text into audio. Far from fluent, but it reads aloud in a characterful voice, doing its earnest best. Best with short text.
- Endpoint
POST /api/v1/tts- I/O
- text->audio
Demo
How to use
Send text, get an audio file (WAV/MP3) back.
Request
curl https://ponkotsu-lab.net/api/v1/tts \
-H "Content-Type: application/json" \
-d '{"text": "Good morning"}' \
--output voice.wav| Field | Type | Required | Description |
|---|---|---|---|
text | string | ✔ | Text to read aloud |
speed | number | Speaking speed (default: 1.0) |
Response
- Binary with
Content-Type: audio/wav.
Limitations (the ponkotsu bits)
- Long text takes a while to synthesize. Short sentences recommended.
- Japanese speech synthesis (powered by OpenJTalk).