Elamur Research⌕
☰
home / papers / 2026 / Qwen3-TTS Technical Report

Qwen3-TTS Technical Report

Paper · 2026 · Hangrui Hu, Xinfa Zhu, Ting He et al.

Summary

Qwen3-TTS is the Qwen team's first text-to-speech family. It aims for synthesis that is multilingual, controllable through natural-language voice descriptions, able to clone a voice from roughly three seconds of audio, and fast enough for streaming. The models use a Qwen3 LM backbone in a dual-track setup: text and speech tokens are combined at each step, so audio is predicted as soon as text arrives.

Two tokenizers are provided:

  • Qwen-TTS-Tokenizer-25Hz is a 25 Hz single-codebook codec built from Qwen2-Audio. It is decoded by a block-wise DiT with flow matching plus BigVGAN.
  • Qwen-TTS-Tokenizer-12Hz is a 12.5 Hz codec with one semantic layer guided by WavLM and 15 RVQ acoustic layers. A small causal ConvNet decodes it, and an MTP module predicts the residual codebooks.

Training covers over 5 million hours in 10 languages. Pre-training runs in three stages (general, high-quality, long-context) and is followed by post-training. The authors report the lowest WER on Seed-TTS zero-shot cloning, better speaker similarity than MiniMax and ElevenLabs, and stable speech beyond 10 minutes. All models and tokenizers are released under Apache 2.0.

Key points

  • The family has 10 checkpoints (0.6B and 1.7B; Base, CustomVoice, VoiceDesign/VoiceEditing variants) for the 12Hz and 25Hz tokenizers. First-packet latency is 97 ms for the 0.6B model and 101 ms for the 1.7B model.
  • The long-context pre-training stage raises the maximum sequence length from 8,192 to 32,768 tokens and upsamples long recordings.
  • In target-speaker fine-tuning (Aiden Voice), the model beats GPT-4o-Audio-Preview on WER in 7 of 10 languages, for example Japanese 3.88 vs. 5.00 and Korean 1.74 vs. 2.76.
  • On an internal long-form set (100 zh/en texts of 200–2000 words), Qwen3-TTS-25Hz-1.7B has the lowest WER against Higgs-Audio-v2, VibeVoice and VoxCPM. The exact figures differ between the paper's text and Table 10.
  • The supported languages are zh, en, de, it, pt, es, ja, ko, fr and ru. Ukrainian is not included.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Fields, topics and concepts

📚 Shelves 7The field of science this work belongs to (OpenAlex fields).Computer Science · 16,667 worksArtificial Intelligence · 16,194 worksArts and Humanities · 16,039 worksLanguage and Linguistics · 15,891 worksSignal Processing · 452 worksMedicine · 241 worksPhysiology · 29 works

Cite

Cite as Hangrui Hu, Xinfa Zhu, Ting He, Dake Guo, Bin Zhang, Xiong Wang, Zhifang Guo, Ziyue Jiang, Hongkun Hao, Zishan Guo, Xinyu Zhang, Pei Zhang, Baosong Yang, Jin Xu, Jingren Zhou and Junyang Lin. 2026. Qwen3-TTS Technical Report. arXiv:2601.15621.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.48550/arxiv.2601.15621 · arXiv 2601.15621

@misc{hu2026qwen3,
  title = {{Qwen3-TTS Technical Report}},
  author = {Hangrui Hu, Xinfa Zhu, Ting He, Dake Guo, Bin Zhang, Xiong Wang, Zhifang Guo, Ziyue Jiang, Hongkun Hao, Zishan Guo, Xinyu Zhang, Pei Zhang, Baosong Yang, Jin Xu, Jingren Zhou and Junyang Lin},
  year = {2026},
  doi = {10.48550/arxiv.2601.15621},
  eprint = {2601.15621},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.48550/arxiv.2601.15621}
}

References

43 references, 14 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Philip Anastassiou, Jiawei Chen, Jitong Chen, Yuanzhe Chen, Zhuo Chen, Ziyi Chen, Jian Cong, Lelai Deng, Chuang Ding, Lu Gao, et al. Seed-tts: A family of high-quality versatile speech generation models. arXiv preprint arXiv:2406.02430, 2024.
  2. Boson AI. Higgs Audio V2: Redefining Expressiveness in Audio Generation. https://github.com/boson -ai/higgs-audio, 2025. GitHub repository. Release blog available at https://www.boson.ai/blog/hi ggs-audio-v2. ↗
  3. Sanyuan Chen, Chengyi Wang, Zhengyang Chen, Yu Wu, Shujie Liu, Zhuo Chen, Jinyu Li, Naoyuki Kanda, Takuya Yoshioka, Xiong Xiao, Jian Wu, Long Zhou, Shuo Ren, Yanmin Qian, Yao Qian, Jian Wu, Michael Zeng, Xiangzhan Yu, and Furu Wei. Wavlm: Large-scale self-supervised pre-training for full stack speec… find
  4. Yushen Chen, Zhikang Niu, Ziyang Ma, Keqi Deng, Chunhui Wang, Jian Zhao, Kai Yu, and Xie Chen. F5-tts: A fairytaler that fakes fluent and faithful speech with flow matching. arXiv preprint arXiv:2410.06885, 2024.
  5. Alexandre Défossez, Jade Copet, Gabriel Synnaeve, and Yossi Adi. High fidelity neural audio compression. arXiv:2210.13438, 2022.
  6. Alexandre Défossez, Laurent Mazaré, Manu Orsini, Amélie Royer, Patrick Pérez, Hervé Jégou, Edouard Grave, and Neil Zeghidour. Moshi: a speech-text foundation model for real-time dialogue. arXiv preprint arXiv:2410.00037, 2024.
  7. Zhihao Du, Qian Chen, Shiliang Zhang, Kai Hu, Heng Lu, Yexin Yang, Hangrui Hu, Siqi Zheng, Yue Gu, Ziyang Ma, et al. Cosyvoice: A scalable multilingual zero-shot text-to-speech synthesizer based on supervised semantic tokens. arXiv preprint arXiv:2407.05407, 2024a. find
  8. Zhihao Du, Yuxuan Wang, Qian Chen, Xian Shi, Xiang Lv, Tianyu Zhao, Zhifu Gao, Yexin Yang, Changfeng Gao, Hui Wang, et al. Cosyvoice 2: Scalable streaming speech synthesis with large language models. arXiv preprint arXiv:2412.10117, 2024b. find
  9. Zhihao Du, Changfeng Gao, Yuxuan Wang, Fan Yu, Tianyu Zhao, Hao Wang, Xiang Lv, Hui Wang, Chongjia Ni, Xian Shi, et al. Cosyvoice 3: Towards in-the-wild speech generation via scaling-up and post-training. arXiv preprint arXiv:2505.17589, 2025. — [arXiv:2505.17589](https://arxiv.org/abs/2505.17589) find
  10. Sefik Emre Eskimez, Xiaofei Wang, Manthan Thakker, Canrun Li, Chung-Hsien Tsai, Zhen Xiao, Hemin Yang, Zirun Zhu, Min Tang, Xu Tan, et al. E2 tts: Embarrassingly easy fully non-autoregressive zero-shot tts. In 2024 IEEE Spoken Language Technology Workshop (SLT), pp. 682–689. find
  11. Yitian Gong, Luozhijie Jin, Ruifan Deng, Dong Zhang, Xin Zhang, Qinyuan Cheng, Zhaoye Fei, Shimin Li, and Xipeng Qiu. Xy-tokenizer: Mitigating the semantic-acoustic conflict in low-bitrate speech codecs. CoRR, abs/2506.23325, 2025.
  12. Dake Guo, Jixun Yao, Linhan Ma, He Wang, and Lei Xie. Streamflow: Streaming flow matching with block-wise guided attention mask for speech token decoding. CoRR, abs/2506.23986, 2025. — [arXiv:2506.23986](https://arxiv.org/abs/2506.23986) find
  13. Zhifang Guo, Yichong Leng, Yihan Wu, Sheng Zhao, and Xu Tan. Prompttts: Controllable text-to-speech with text descriptions. In IEEE International Conference on Acoustics, Speech and Signal Processing ICASSP 2023, Rhodes Island, Greece, June 4-10, 2023, pp. 1–5. find
  14. IEEE, 2023. \#\# 1 Alphabetical order. † Corresponding Authors. find
  15. Jingbin Hu, Huakang Chen, Linhan Ma, Dake Guo, Qirui Zhan, Wenhao Li, Haoyu Zhang, Kangxiang Xia, Ziyu Zhang, Wenjie Tian, et al. Voicesculptor: Your voice, designed by you. arXiv preprint arXiv:2601.10629, 2026. — [arXiv:2601.10629](https://arxiv.org/abs/2601.10629) find
  16. Kexin Huang, Qian Tu, Liwei Fan, Chenchen Yang, Dong Zhang, Shimin Li, Zhaoye Fei, Qinyuan Cheng, and Xipeng Qiu. Instructttseval: Benchmarking complex natural-language instruction following in text-to-speech systems. CoRR, abs/2506.16381, 2025. find
  17. Shengpeng Ji, Qian Chen, Wen Wang, Jialong Zuo, Minghui Fang, Ziyue Jiang, Hai Huang, Zehan Wang, Xize Cheng, Siqi Zheng, et al. Controlspeech: Towards simultaneous and independent zero-shot speaker cloning and zero-shot language style control. In Proceedings of the 63rd Annual Meeting of the Associ…
  18. Zeqian Ju, Yuancheng Wang, Kai Shen, Xu Tan, Detai Xin, Dongchao Yang, Yanqing Liu, Yichong Leng, Kaitao Song, Siliang Tang, et al. Naturalspeech 3: Zero-shot speech synthesis with factorized codec and diffusion models. arXiv preprint arXiv:2403.03100, 2024. — [arXiv:2403.03100](https://arxiv.org/abs/2403.03100) find
  19. Rithesh Kumar, Prem Seetharaman, Alejandro Luebs, Ishaan Kumar, and Kundan Kumar. High-fidelity audio compression with improved rvqgan. Advances in Neural Information Processing Systems, 36: 27980–27993, 2023. find
  20. Sang-gil Lee, Wei Ping, Boris Ginsburg, Bryan Catanzaro, and Sungroh Yoon. Bigvgan: A universal neural vocoder with large-scale training. In ICLR 2023. find
  21. Yichong Leng, Zhifang Guo, Kai Shen, Zeqian Ju, Xu Tan, Eric Liu, Yufei Liu, Dongchao Yang, Kaitao Song, Lei He, et al. Prompttts 2: Describing and generating voices with text prompt. In The Twelfth International Conference on Learning Representations. Shijia Liao, Yuxuan Wang, Tianyu Li, Yifan Chen… — [arXiv:2411.01156](https://arxiv.org/abs/2411.01156) find
  22. Yaron Lipman, Ricky T. Q. Chen, Heli Ben-Hamu, Maximilian Nickel, and Matthew Le. Flow matching for generative modeling. In ICLR 2023.
  23. Guanghou Liu, Yongmao Zhang, Yi Lei, Yunlin Chen, Rui Wang, Lei Xie, and Zhifei Li. Promptstyle: Controllable style transfer for text-to-speech with natural language descriptions. In Naomi Harte, Julie Carson-Berndsen, and Gareth Jones (eds.), 24th Annual Conference of the International Speech Commu… find
  24. Dan Lyth and Simon King. Natural language guidance of high-fidelity text-to-speech with synthetic annotations. arXiv preprint arXiv:2402.01912, 2024. — [arXiv:2402.01912](https://arxiv.org/abs/2402.01912) find
  25. Zhiliang Peng, Jianwei Yu, Wenhui Wang, Yaoyao Chang, Yutao Sun, Li Dong, Yi Zhu, Weijiang Xu, Hangbo Bao, Zehua Wang, Shaohan Huang, Yan Xia, and Furu Wei. Vibevoice technical report. CoRR, abs/2508.19205, 2025. — [arXiv:2508.19205](https://arxiv.org/abs/2508.19205) find
  26. Rafael Rafailov, Archit Sharma, Eric Mitchell, Christopher D. Manning, Stefano Ermon, and Chelsea Finn. Direct preference optimization: Your language model is secretly a reward model. In NeurIPS, 2023.
  27. Kai Shen, Zeqian Ju, Xu Tan, Yanqing Liu, Yichong Leng, Lei He, Tao Qin, Sheng Zhao, and Jiang Bian. Naturalspeech 2: Latent diffusion models are natural and zero-shot speech and singing synthesizers. arXiv preprint arXiv:2304.09116, 2023. — [arXiv:2304.09116](https://arxiv.org/abs/2304.09116) find
  28. Chengyi Wang, Sanyuan Chen, Yu Wu, Ziqiang Zhang, Long Zhou, Shujie Liu, Zhuo Chen, Yanqing Liu, Huaming Wang, Jinyu Li, et al. Neural codec language models are zero-shot text to speech synthesizers. arXiv preprint arXiv:2301.02111, 2023.
  29. Xinsheng Wang, Mingqi Jiang, Ziyang Ma, Ziyu Zhang, Songxiang Liu, Linqin Li, Zheng Liang, Qixi Zheng, Rui Wang, Xiaoqin Feng, et al. Spark-tts: An efficient llm-based text-to-speech model with single-stream decoupled speech tokens. arXiv preprint arXiv:2503.01710, 2025. — [arXiv:2503.01710](https://arxiv.org/abs/2503.01710) find
  30. Yuancheng Wang, Haoyue Zhan, Liwei Liu, Ruihong Zeng, Haotian Guo, Jiachen Zheng, Qiang Zhang, Xueyao Zhang, Shunsi Zhang, and Zhizheng Wu. Maskgct: Zero-shot text-to-speech with masked generative codec transformer. arXiv preprint arXiv:2409.00750, 2024. — [arXiv:2409.00750](https://arxiv.org/abs/2409.00750) find
  31. Kangxiang Xia, Xinfa Zhu, Jixun Yao, Wenjie Tian, Wenhao Li, and Lei Xie. Kall-e: Autoregressive speech synthesis with next-distribution prediction. CoRR, abs/2412.16846, 2024. — [arXiv:2412.16846](https://arxiv.org/abs/2412.16846) find
  32. Kun Xie, Feiyu Shen, Junjie Li, Fenglong Xie, Xu Tang, and Yao Hu. Fireredtts-2: Towards long conversational speech generation for podcast and chatbot. CoRR, abs/2509.02020, 2025. — [arXiv:2509.02020](https://arxiv.org/abs/2509.02020) find
  33. Jin Xu, Zhifang Guo, Jinzheng He, Hangrui Hu, Ting He, Shuai Bai, Keqin Chen, Jialin Wang, Yang Fan, Kai Dang, et al. Qwen2. 5-omni technical report. arXiv preprint arXiv:2503.20215, 2025.
  34. Dongchao Yang, Jinchuan Tian, Xu Tan, Rongjie Huang, Songxiang Liu, Xuankai Chang, Jiatong Shi, Sheng Zhao, Jiang Bian, Xixin Wu, et al. Uniaudio: An audio foundation model toward universal audio generation. arXiv preprint arXiv:2310.00704, 2023. — [arXiv:2310.00704](https://arxiv.org/abs/2310.00704) find
  35. Zhen Ye, Peiwen Sun, Jiahe Lei, Hongzhan Lin, Xu Tan, Zheqi Dai, Qiuqiang Kong, Jianyi Chen, Jiahao Pan, Qifeng Liu, Yike Guo, and Wei Xue. Codec does matter: Exploring the semantic shortcoming of codec for audio language model. In AAAI-25, Sponsored by the Association for the Advancement of Artific… find
  36. Zhen Ye, Xinfa Zhu, Chi-Min Chan, Xinsheng Wang, Xu Tan, Jiahe Lei, Yi Peng, Haohe Liu, Yizhu Jin, Zheqi Dai, Hongzhan Lin, Jianyi Chen, Xingjian Du, Liumeng Xue, Yunlin Chen, Zhifei Li, Lei Xie, Qiuqiang Kong, Yike Guo, and Wei Xue. Llasa: Scaling train-time and inference-time compute for llama-bas… — [arXiv:2502.04128](https://arxiv.org/abs/2502.04128) find
  37. Neil Zeghidour, Alejandro Luebs, Ahmed Omran, Jan Skoglund, and Marco Tagliasacchi. Soundstream: An end-to-end neural audio codec. IEEE ACM Trans. Audio Speech Lang. Process., 2022.
  38. Bowen Zhang, Congchao Guo, Geng Yang, Hang Yu, Haozhe Zhang, Heidi Lei, Jialong Mai, Junjie Yan, Kaiyue Yang, Mingqi Yang, et al. Minimax-speech: Intrinsic zero-shot text-to-speech with a learnable speaker encoder. arXiv preprint arXiv:2505.07916, 2025a. — [arXiv:2505.07916](https://arxiv.org/abs/2505.07916) find
  39. Dong Zhang, Gang Wang, Jinlong Xue, Kai Fang, Liang Zhao, Rui Ma, Shuhuai Ren, Shuo Liu, Tao Guo, Weiji Zhuang, et al. Mimo-audio: Audio language models are few-shot learners. arXiv preprint arXiv:2512.23808, 2025b.
  40. Xin Zhang, Dong Zhang, Shimin Li, Yaqian Zhou, and Xipeng Qiu. Speechtokenizer: Unified speech tokenizer for speech large language models. CoRR, abs/2308.16692, 2023a.
  41. Xin Zhang, Dong Zhang, Shimin Li, Yaqian Zhou, and Xipeng Qiu. Speechtokenizer: Unified speech tokenizer for speech large language models. arXiv preprint arXiv:2308.16692, 2023b.
  42. Yixuan Zhou, Xiaoyu Qin, Zeyu Jin, Shuoyi Zhou, Shun Lei, Songtao Zhou, Zhiyong Wu, and Jia Jia. Voxinstruct: Expressive human instruction-to-speech generation with unified multilingual codec language modelling. In Jianfei Cai, Mohan S. Kankanhalli, Balakrishnan Prabhakaran, Susanne Boll, Ramanathan… find
  43. Yixuan Zhou, Guoyang Zeng, Xin Liu, Xiang Li, Renjie Yu, Ziyang Wang, Runchuan Ye, Weiyue Sun, Jiancheng Gui, Kehan Li, Zhiyong Wu, and Zhiyuan Liu. Voxcpm: Tokenizer-free TTS for context-aware speech generation and true-to-life voice cloning. CoRR, abs/2509.24650, 2025. — [arXiv:2509.24650](https://arxiv.org/abs/2509.24650) find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/hu-2026-qwen3-tts-technical-report/