mlmihjaz/dhivehi-tts-api

Dhivehi TTS API

Real-time Dhivehi Text-to-Speech API using Facebook MMS-TTS.

Endpoints

POST /synthesize

Send text, get full WAV audio back.

{
  "text": "ދިވެހިރާއްޖެ",
  "voice": "female",
  "speed": 1.0
}

POST /stream

Stream audio chunk by chunk per sentence.

GET /status

Check if model is loaded.

Voices

• —female — Aminath (default, original model voice)
• —male — Ibrahim (pitch shifted down)