Qwen3-TTS - OpenAI-Compatible TTS API

Text-to-Speech Demo

Enter Text

Voice

Model

Language

Audio Format

Voice Style Instructions (optional)

Speed: 1.0x

Advanced text normalization options

Enable normalization Unit normalization URL normalization Email normalization Phone normalization Optional pluralization normalization Replace remaining symbols

🎙️ Voice Cloning

Checking voice cloning capabilities...

Clone Mode

Reference Audio Upload a 5-10 second audio file (WAV, MP3, etc.)

Reference Audio Preview

Reference Text (Transcript) Required for ICL mode. Type exactly what is said in the reference audio.

Text to Synthesize

Language

Audio Format

Speed: 1.0x

Generated Audio

Backend Status

Loading backend information...

API Endpoints

📖 /docs Swagger UI Documentation 📚 /redoc ReDoc Documentation 🤖 /v1/models List Available Models 🎤 /v1/voices List Available Voices

Python Example

                    
from openai import OpenAI

# Create client pointing to local server
client = OpenAI(
    base_url="http://localhost:8880/v1",
    api_key="not-needed"
)

# Generate speech
response = client.audio.speech.create(
    model="qwen3-tts",
    voice="Vivian",
    input="Hello! This is Qwen3-TTS."
)

# Save to file
response.stream_to_file("output.mp3")
                    
                

Acknowledgments

This project is built on top of the incredible Qwen3-TTS model developed by the Qwen Team at Alibaba Cloud. We are deeply grateful for their outstanding work in advancing the state-of-the-art in text-to-speech technology.

The Qwen3-TTS model represents a significant advancement in neural speech synthesis, offering:

🎯 High-quality, natural-sounding voice synthesis
🌍 Support for 10+ languages
🎨 Advanced voice cloning and design capabilities
⚡ Efficient inference for production deployments

📦 Qwen3-TTS GitHub → 🤗 Hugging Face Models → 📑 Research Paper →