Qwen3-TTS Documentation

A powerful Python SDK for text-to-speech generation with voice cloning, voice design, and ultra-low latency streaming supporting 10 major languages.

Get Started API Reference

Quick Start

Get up and running with Qwen3-TTS in minutes

Install Qwen3-TTS using pip in a fresh Python environment: ```bash theme={null} pip install -U qwen-tts ``` We recommend using Python 3.12 with a clean conda or virtual environment to avoid dependency conflicts. Import and initialize a Qwen3-TTS model: ```python theme={null} import torch from qwen_tts import Qwen3TTSModel model = Qwen3TTSModel.from_pretrained( "Qwen/Qwen3-TTS-12Hz-1.7B-CustomVoice", device_map="cuda:0", dtype=torch.bfloat16, ) ``` Generate natural-sounding speech from text: ```python theme={null} import soundfile as sf wavs, sr = model.generate_custom_voice( text="Hello, welcome to Qwen3-TTS!", language="English", speaker="Ryan", ) sf.write("output.wav", wavs[0], sr) ``` The model automatically downloads from Hugging Face on first use. For offline environments, see the [Installation](/installation) guide.

Explore by Feature

Discover what you can build with Qwen3-TTS

Generate speech with 9 premium preset voices covering multiple languages and dialects Create custom voices from natural language descriptions with instruction-based control Clone any voice in just 3 seconds from a reference audio sample Ultra-low latency streaming with 97ms end-to-end synthesis for real-time interactions Process multiple text inputs efficiently with batched inference Customize models for your specific use case with fine-tuning

Key Features

What makes Qwen3-TTS powerful

10 Languages

Support for Chinese, English, Japanese, Korean, German, French, Russian, Portuguese, Spanish, and Italian with multilingual and cross-lingual capabilities.