metadata
title: SuperTonic3 TTS API
emoji: 🎤
colorFrom: blue
colorTo: purple
sdk: docker
pinned: false
license: mit
SuperTonic3 TTS API
OpenAI-compatible Text-to-Speech API server powered by supertonic3 (ONNX-based). Drop-in replacement for OpenAI's TTS API with 31 languages and 10 built-in voices.
Key Features:
- OpenAI API Compatible - Works with any OpenAI TTS client
- 31 Languages - Single model supports 31 languages
- 10 Built-in Voices - M1-M5 (male), F1-F5 (female)
- CPU Optimized - No GPU required (ONNX runtime)
- 44100 Hz Output - High-quality audio
- Docker Ready - One-command deployment
Quick Start
Docker
docker compose up -d
Server available at http://localhost:7860
Python
python -m venv venv
source venv/bin/activate
pip install -r requirements.txt
python server.py
API Usage
Generate Speech
curl -X POST http://localhost:7860/v1/audio/speech \
-H "Content-Type: application/json" \
-d '{
"model": "tts-1",
"input": "Hello world!",
"voice": "M1"
}' \
--output speech.mp3
Python Client
from openai import OpenAI
client = OpenAI(
base_url="http://localhost:7860/v1",
api_key="not-needed"
)
response = client.audio.speech.create(
model="tts-1",
voice="M1",
input="Hello world!"
)
response.stream_to_file("output.mp3")
Speech Parameters
| Parameter | Type | Required | Default | Description |
|---|---|---|---|---|
input |
string | Yes | - | Text to synthesize |
voice |
string | No | M1 |
Voice: M1-M5, F1-F5 |
response_format |
string | No | mp3 |
Output: mp3, wav, flac |
lang |
string | No | en |
Language code (31 supported) |
stream |
boolean | No | false |
Enable streaming |
Languages
31 supported languages: en, zh, ja, ko, fr, de, es, it, pt, ru, ar, hi, bn, id, ms, th, vi, tl, tr, fa, pl, nl, sv, da, fi, cs, ro, hu, el, he, uk
Configuration
| Environment Variable | Default | Description |
|---|---|---|
SUPERTONIC3_HOST |
0.0.0.0 |
Bind address |
SUPERTONIC3_PORT |
7860 |
Port |
SUPERTONIC3_VOICE |
M1 |
Default voice |
SUPERTONIC3_LOG_LEVEL |
INFO |
Log level |