In this report, we present the Qwen3-TTS series, a family of advanced multilingual, controllable, robust, and streaming text-to-speech models. Qwen3-TTS supports state-of-the-art 3-second voice cloning and description-based control, allowing both the creation of entirely novel voices and fine-grained manipulation over the output speech. Trained on over 5 million hours of speech data spanning 10 languages, Qwen3-TTS adopts a dual-track LM architecture for real-time synthesis, coupled with two speech tokenizers: 1) Qwen-TTS-Tokenizer-25Hz is a single-codebook codec emphasizing semantic content, which offers seamlessly integration with Qwen-Audio and enables streaming waveform reconstruction via a block-wise DiT. 2) Qwen-TTS-Tokenizer-12Hz achieves extreme bitrate reduction and ultra-low-latency streaming, enabling immediate first-packet emission ($97\,\mathrm{ms}$) through its 12.5 Hz, 16-layer multi-codebook design and a lightweight causal ConvNet. Extensive experiments indicate state-of-the-art performance across diverse objective and subjective benchmark (e.g., TTS multilingual test set, InstructTTSEval, and our long speech test set). To facilitate community research and development, we release both tokenizers and models under the Apache 2.0 license.

翻译：在本报告中，我们介绍了Qwen3-TTS系列模型，这是一组先进的多语言、可控、鲁棒且支持流式合成的文本转语音模型。Qwen3-TTS支持业界领先的3秒语音克隆和基于描述的语音控制，既能生成全新的语音，也能对输出语音进行细粒度操控。该模型基于超过500万小时、涵盖10种语言的语音数据进行训练，采用双轨语言模型架构实现实时合成，并配备两种语音分词器：1) Qwen-TTS-Tokenizer-25Hz是一种强调语义内容的单码本编解码器，可与Qwen-Audio无缝集成，并通过分块DiT实现流式波形重建。2) Qwen-TTS-Tokenizer-12Hz实现了极致的比特率压缩和超低延迟流式合成，其12.5 Hz、16层多码本设计结合轻量级因果卷积网络，可实现首个数据包的即时发射（延迟为$97\,\mathrm{ms}$）。大量实验表明，该模型在多样化的客观与主观评测基准（如TTS多语言测试集、InstructTTSEval以及我们的长语音测试集）上均达到了业界领先的性能。为促进社区研究与开发，我们依据Apache 2.0许可证开源了所有分词器与模型。