Elamur Research⌕
☰
home / papers / 2025 / Hunyuan-MT Technical Report

Hunyuan-MT Technical Report

Paper · 2025 · Mao Zheng, Zheng Li, Bingxin Qu et al.

Summary

Open-source LLM translators still trail proprietary systems, and translation between Mandarin and China's minority languages gets little attention. Tencent's Hunyuan Team addresses both problems with Hunyuan-MT-7B, a 7B-parameter model that translates in both directions among 33 languages. It specifically targets Mandarin paired with Kazakh, Uyghur, Mongolian and Tibetan, as well as dialects.

Training proceeds in stages:

  • general multilingual pre-training on quality-scored and topic-balanced data;
  • MT-oriented pre-training on cleaned monolingual and parallel corpora;
  • SFT, then reinforcement learning.

A further weak-to-strong RL stage produces Hunyuan-MT-Chimera-7B. Instead of chain-of-thought reasoning, this companion model takes several candidate translations from the base model, generated with different settings, and fuses them into one better output.

The models beat translation-specialised systems of similar size and many much larger LLMs, with the clearest advantage on minority-language directions. They placed first in 30 of 31 language pairs at WMT2025 General MT. The weights are public, which gives low-resource MT research a strong and inexpensive foundation.

Key points

  • General pre-training data: the multilingual portion has 1.3 trillion tokens covering 112 languages and dialects besides Chinese and English. A quality model scores Knowledge Value, Authenticity and Writing Style on a 0–2 scale, and industry and theme tags (24 categories each) balance the topic mix.
  • MT-oriented pre-training data:
    • Monolingual text comes from mC4 and OSCAR, cleaned with fastText language ID, minLSH deduplication and KenLM perplexity filtering.
    • Parallel text comes from OPUS and ParaCrawl, filtered with CometKiwi.
    • The data mixture is chosen in the style of RegMix, and 20% of the original pre-training corpus is replayed to limit forgetting.
  • WMT2025 General MT: ranked first in 30 of 31 language pairs, spanning high-resource languages such as Chinese and Japanese and lower-resource ones such as Icelandic and Marathi.
  • Human evaluation (ZH⇔EN, multi-domain challenge set): Hunyuan-MT-7B averaged 3.189, close to Gemini-2.5-Pro (3.223) and DeepSeek-V3-0324 (3.219). Google-Translator scored 2.344.
  • Chain-of-thought finding: rewarding only the final translation yielded generic, empty reasoning and no quality gain. Rewarding both the reasoning and the output, following the authors' TAT-R1 approach, did improve quality.

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 4The 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 works

🧭 Topics 2Research subjects and languages the work is about.Language models & pretraining · 3,320 worksMachine translation · 1,087 works

Cite

Cite as Mao Zheng, Zheng Li, Bingxin Qu, Mingyang Song, Yang Du, Mingrui Sun and Di Wang. 2025. Hunyuan-MT Technical Report. arXiv:2509.05209.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.48550/arxiv.2509.05209 · arXiv 2509.05209

@misc{zheng2025hunyuan,
  title = {{Hunyuan-MT Technical Report}},
  author = {Mao Zheng, Zheng Li, Bingxin Qu, Mingyang Song, Yang Du, Mingrui Sun and Di Wang},
  year = {2025},
  doi = {10.48550/arxiv.2509.05209},
  eprint = {2509.05209},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.48550/arxiv.2509.05209}
}

References

60 references, 31 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Rishabh Agarwal, Avi Singh, Lei Zhang, Bernd Bohnet, Luis Rosias, Stephanie C. Y. Chan, Biao Zhang, Ankesh Anand, Zaheer Abbas, Azade Nova, John D. Co-Reyes, Eric Chu, Feryal M. P. Behbahani, Aleksandra Faust, and Hugo Larochelle. Many-shot in-context learning. In Amir Globersons, Lester Mackey, Dan…
  2. AI@Meta. Llama-3.1-8b-instruct. https://huggingface.co/meta-llama/Llama-4-Scout-17B-16E-Instr uct, 2025a. Accessed: 2025-08-26. ↗
  3. AI@Meta. Llama-4-scout-17b-16e-instruct. https://huggingface.co/meta-llama/Llama-4-Scout-17B-1 6E-Instruct, 2025b. Accessed: 2025-08-26. ↗
  4. Anthropic. Introducing claude 4. https://www.anthropic.com/news/claude-4, 2025. Accessed: 2025-08-26. ↗
  5. Dzmitry Bahdanau, Kyunghyun Cho, and Yoshua Bengio. Neural machine translation by jointly learning to align and translate. In Yoshua Bengio and Yann LeCun (eds.), 3rd International Conference on Learning Representations, ICLR 2015, San Diego, CA, USA, May 7-9, 2015, Conference Track Proceedings, 201…
  6. Peter F. Brown, John Cocke, Stephen Della Pietra, Vincent J. Della Pietra, Frederick Jelinek, John D. Lafferty, Robert L. Mercer, and Paul S. Roossin. A statistical approach to machine translation. Comput. Linguistics, 16(2):79–85, 1990.
  7. Peter F. Brown, Stephen Della Pietra, Vincent J. Della Pietra, and Robert L. Mercer. The mathematics of statistical machine translation: Parameter estimation. Comput. Linguistics, 19(2):263–311, 1993. find
  8. Christian Buck and Philipp Koehn. Findings of the wmt 2016 bilingual document alignment shared task. In Proceedings of the First Conference on Machine Translation, pp. 554–563, Berlin, Germany, August 2016. Association for Computational Linguistics. URL http://www.aclweb.org/anthology/W/W16/W16-2347… ↗
  9. Federico Cassano, John Gouwar, Daniel Nguyen, Sydney Nguyen, Luna Phipps-Costin, Donald Pinckney, Ming-Ho Yee, Yangtian Zi, Carolyn Jane Anderson, Molly Q. Feldman, Arjun Guha, Michael Greenberg, and Abhinav Jangda. Multipl-e: A scalable and polyglot approach to benchmarking neural code generation.… — [doi:10.1109/tse.2023.3267446](https://doi.org/10.1109/tse.2023.3267446) find
  10. Shanbo Cheng, Yu Bao, Qian Cao, Luyang Huang, Liyan Kang, Zhicheng Liu, Yu Lu, Wenhao Zhu, Jingwen Chen, Zhichao Huang, Tao Li, Yifu Li, Huiying Lin, Sitong Liu, Ningxin Peng, Shuaijie She, Lu Xu, Nuo Xu, Sen Yang, Runsheng Yu, Yiming Yu, Liehao Zou, Hang Li, Lu Lu, Yuxuan Wang, and Yonghui Wu. Seed… — [arXiv:2507.13618](https://arxiv.org/abs/2507.13618) find
  11. Karl Cobbe, Vineet Kosaraju, Mohammad Bavarian, Mark Chen, Heewoo Jun, Lukasz Kaiser, Matthias Plappert, Jerry Tworek, Jacob Hilton, Reiichiro Nakano, Christopher Hesse, and John Schulman. Training verifiers to solve math word problems. CoRR, abs/2110.14168, 2021. URL https://arxiv.or g/abs/2110.141… ↗
  12. Menglong Cui, Pengzhi Gao, Wei Liu, Jian Luan, and Bin Wang. Multilingual machine translation with open large language models at practical scale: An empirical study, 2025. URL https://arxiv.org/ab s/2502.02481. ↗
  13. DeepMind. We’re expanding our gemini 2.5 family of models. https://blog.google/products/gemini/gemini-2-5-model-family-expands/, 2025. Accessed: 2025-08-26. ↗
  14. DeepSeek-AI. Deepseek-v3 technical report, 2024. URL https://arxiv.org/abs/2412.19437. ↗
  15. DeepSeek-AI. Deepseek-r1: Incentivizing reasoning capability in llms via reinforcement learning, 2025. URL https://arxiv.org/abs/2501.12948. ↗
  16. Daniel Deutsch, Eleftheria Briakou, Isaac Caswell, Mara Finkelstein, Rebecca Galor, Juraj Juraska, Geza Kovacs, Alison Lui, Ricardo Rei, Jason Riesa, Shruti Rijhwani, Parker Riley, Elizabeth Salesky, Firas Trabelsi, Stephanie Winkler, Biao Zhang, and Markus Freitag. WMT24++: Expanding the Language C… — WMT24++: Expanding the Language Coverage of WMT24 to 55 Languages & Dialects
  17. Alex Gu, Baptiste Rozière, Hugh James Leather, Armando Solar-Lezama, Gabriel Synnaeve, and Sida Wang. Cruxeval: A benchmark for code reasoning, understanding and execution. In Forty-first International Conference on Machine Learning, ICML 2024, Vienna, Austria, July 21-27, 2024.
  18. OpenReview.net, 2024. URL https://openreview.net/forum?id=Ffpg52swvg. ↗
  19. Nuno M. Guerreiro, Ricardo Rei, Daan van Stigt, Luisa Coheur, Pierre Colombo, and André F. T. Martins. xcomet: Transparent machine translation evaluation through fine-grained error detection, 2023. URL https://arxiv.org/abs/2310.10482. ↗
  20. Dan Hendrycks, Collin Burns, Steven Basart, Andy Zou, Mantas Mazeika, Dawn Song, and Jacob Steinhardt. Measuring massive multitask language understanding. In 9th International Conference on Learning Representations, ICLR 2021, Virtual Event, Austria, May 3-7, 2021.
  21. OpenReview.net, 2021. URL https://openreview.net/forum?id=d7KBjmI3GmQ. ↗
  22. Bojie Hu, Ambyer Han, Zheyang Zhang, Shen Huang, and Qi Ju. Tencent minority-mandarin translation system. In Shujian Huang and Kevin Knight (eds.), Machine Translation, pp. 93–104, Singapore, 2019. Springer Singapore. ISBN 978-981-15-1721-1. find
  23. Wenxiang Jiao, Wenxuan Wang, Jen-tse Huang, Xing Wang, and Zhaopeng Tu. Is chatgpt A good translator? A preliminary study. CoRR, abs/2301.08745, 2023. doi: 10.48550/ARXIV.2301.08745. URL https://doi.org/10.48550/arXiv.2301.08745. ↗
  24. Tom Kocmi and Christian Federmann. Large language models are state-of-the-art evaluators of translation quality. In Proceedings of the 24th Annual Conference of the European Association for Machine Translation, pp. 193–203, Tampere, Finland, June 2023. European Association for Machine Translation. U… find
  25. Tom Kocmi, Eleftherios Avramidis, Rachel Bawden, Ondrej Bojar, Anton Dvorkovich, Christian Federmann, Mark Fishel, Markus Freitag, Thamme Gowda, Roman Grundkiewicz, Barry Haddow, Marzena Karpinska, Philipp Koehn, Benjamin Marie, Christof Monz, Kenton Murray, Masaaki Nagata, Martin Popel, Maja Popovi… — Findings of the WMT24 General Machine Translation Shared Task: The LLM Era Is Here but MT Is Not Solved Yet
  26. Zheng Li, Mao Zheng, Mingyang Song, and Wenjie Yang. Tat-r1: Terminology-aware translation with reinforcement learning and word alignment, 2025. URL https://arxiv.org/abs/2505.21172. — [arXiv:2505.21172](https://arxiv.org/abs/2505.21172) ↗
  27. Cong Lin and Liz Jackson. Assimilation over protection: rethinking mandarin language assimilation in china. Multicultural Education Review, 13(4):338–361, 2021. doi: 10.1080/2005615X.2021.2006117. — [doi:10.1080/2005615x.2021.2006117](https://doi.org/10.1080/2005615x.2021.2006117) find
  28. Qian Liu, Xiaosen Zheng, Niklas Muennighoff, Guangtao Zeng, Longxu Dou, Tianyu Pang, Jing Jiang, and Min Lin. Regmix: Data mixture as regression for language model pre-training, 2025. URL https://arxiv.org/abs/2407.01492. ↗
  29. Michael Luo, Sijun Tan, Roy Huang, Ameen Patel, Alpay Ariyak, Qingyang Wu, Xiaoxiang Shi, Rachel Xin, Colin Cai, Maurice Weber, Ce Zhang, Li Erran Li, Raluca Ada Popa, and Ion Stoica. Deepcoder: A fully open-source 14b coder at o3-mini level. https://pretty-radio-b75.notion.site/DeepCod er-A-Fully-O… ↗
  30. Niklas Muennighoff, Zitong Yang, Weijia Shi, Xiang Lisa Li, Li Fei-Fei, Hannaneh Hajishirzi, Luke Zettlemoyer, Percy Liang, Emmanuel J. Candès, and Tatsunori Hashimoto. s1: Simple test-time scaling. CoRR, abs/2501.19393, 2025. doi: 10.48550/ARXIV.2501.19393. URL https://doi.org/10.48550/arX iv.2501… ↗
  31. OpenAI. Introducing gpt-4.1 in the api. https://openai.com/index/gpt-4-1/, 2025. Accessed: 2025-08-26. ↗
  32. Pedro Javier Ortiz Su’arez, Benoit Sagot, and Laurent Romary. Asynchronous pipelines for processing huge corpora on medium to low resource infrastructures. Proceedings of the Workshop on Challenges in the Management of Large Corpora (CMLC-7) 2019. Cardiff, 22nd July 2019, pp. 9 – 16, Mannheim, 2019.… — [doi:10.14618/ids-pub-9021](https://doi.org/10.14618/ids-pub-9021) find
  33. Pedro Javier Ortiz Su’arez, Laurent Romary, and Benoit Sagot. A monolingual approach to contextualized word embeddings for mid-resource languages. In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pp. 1703–1714, Online, July 2020. Association for Computation…
  34. Jianhui Pang, Fanghua Ye, Derek Fai Wong, Dian Yu, Shuming Shi, Zhaopeng Tu, and Longyue Wang. Salute the classic: Revisiting challenges of machine translation in the age of large language models. Trans. Assoc. Comput. Linguistics, 13:73–95, 2025. doi: 10.1162/TACL \\ A \\ 00730. URL https://doi.org/…
  35. Kishore Papineni, Salim Roukos, Todd Ward, and Wei-Jing Zhu. Bleu: a method for automatic evaluation of machine translation. In Proceedings of the 40th Annual Meeting of the Association for Computational Linguistics, July 6-12, 2002, Philadelphia, PA, USA, pp. 311–318.
  36. ACL, 2002. doi: 10.3115/1073083.1073135. URL https://aclanthology.org/P02-1040/. ↗
  37. Colin Raffel, Noam Shazeer, Adam Roberts, Katherine Lee, Sharan Narang, Michael Matena, Yanqi Zhou, Wei Li, and Peter J. Liu. Exploring the limits of transfer learning with a unified text-to-text transformer. arXiv e-prints, 2019.
  38. Ricardo Rei, Marcos Treviso, Nuno M. Guerreiro, Chrysoula Zerva, Ana C. Farinha, Christine Maroti, José G. C. de Souza, Taisiya Glushkova, Duarte M. Alves, Alon Lavie, Luisa Coheur, and André F. T. Martins. Cometkiwi: Ist-unbabel 2022 submission for the quality estimation shared task, 2022. URL ht…
  39. Ricardo Rei, Nuno M. Guerreiro, José Pombal, João Alves, Pedro Teixeirinha, Amin Farajian, and André F. T. Martins. Tower+: Bridging generality and translation specialization in multilingual llms, 2025. URL https://arxiv.org/abs/2506.17080. ↗
  40. David Rein, Betty Li Hou, Asa Cooper Stickland, Jackson Petty, Richard Yuanzhe Pang, Julien Dirani, Julian Michael, and Samuel R. Bowman. GPQA: A graduate-level google-proof q&a benchmark. CoRR, abs/2311.12022, 2023. doi: 10.48550/ARXIV.2311.12022. URL https://doi.org/10.48550/arXiv.2311.12022. ↗
  41. Angelika Romanou, Negar Foroutan, Anna Sotnikova, Zeming Chen, Sree Harsha Nelaturu, Shivalika Singh, Rishabh Maheshwary, Micol Altomare, Mohamed A. Haggag, Imanol Schlag, Marzieh Fadaee, Sara Hooker, Antoine Bosselut, Snegha A, Alfonso Amayuelas, Azril Hafizi Amirudin, Viraat Aryabumi, Danylo Boiko… — INCLUDE: Evaluating Multilingual Language Understanding with Regional Knowledge
  42. OpenReview.net, 2025. URL https://openreview.net/forum?id=k3gCieTXeY. ↗
  43. Zhihong Shao, Peiyi Wang, Qihao Zhu, Runxin Xu, Junxiao Song, Mingchuan Zhang, Y. K. Li, Y. Wu, and Daya Guo. Deepseekmath: Pushing the limits of mathematical reasoning in open language models. CoRR, abs/2402.03300, 2024. doi: 10.48550/ARXIV.2402.03300. URL https://doi.org/10.48550/arX iv.2402.03300… ↗
  44. Freda Shi, Mirac Suzgun, Markus Freitag, Xuezhi Wang, Suraj Srivats, Soroush Vosoughi, Hyung Won Chung, Yi Tay, Sebastian Ruder, Denny Zhou, Dipanjan Das, and Jason Wei. Language models are multilingual chain-of-thought reasoners. In The Eleventh International Conference on Learning Representations,…
  45. OpenReview.net, 2023. URL https://openreview.net/forum?id=fR3wGCk-IXp. ↗
  46. Mingyang Song, Mao Zheng, Zheng Li, Wenjie Yang, Xuan Luo, Yue Pan, and Feng Zhang. Fastcurl: Curriculum reinforcement learning with stage-wise context scaling for efficient training r1-like reasoning models, 2025a. URL https://arxiv.org/abs/2503.17287. ↗
  47. Mingyang Song, Mao Zheng, and Xuan Luo. Can many-shot in-context learning help llms as evaluators? A preliminary empirical study. In Owen Rambow, Leo Wanner, Marianna Apidianaki, Hend Al-Khalifa, Barbara Di Eugenio, and Steven Schockaert (eds.), Proceedings of the 31st International Conference on Co…
  48. Ilya Sutskever, Oriol Vinyals, and Quoc V. Le. Sequence to sequence learning with neural networks. In Zoubin Ghahramani, Max Welling, Corinna Cortes, Neil D. Lawrence, and Kilian Q. Weinberger (eds.), Advances in Neural Information Processing Systems 27: Annual Conference on Neural Information Proce…
  49. Gemma Team, Aishwarya Kamath, Johan Ferret, Shreya Pathak, Nino Vieillard, Ramona Merhej, Sarah Perrin, Tatiana Matejovicova, Alexandre Ramé, Morgane Rivière, Louis Rouillard, Thomas Mesnard, Geoffrey Cideron, Jean bastien Grill, Sabela Ramos, Edouard Yvinec, Michelle Casbon, Etienne Pot, Ivo Penc… find
  50. Kimi Team, Yifan Bai, Yiping Bao, Guanduo Chen, Jiahao Chen, Ningxin Chen, Ruijue Chen, Yanru Chen, Yuankun Chen, Yutian Chen, Zhuofu Chen, Jialei Cui, Hao Ding, Mengnan Dong, Angang Du, Chenzhuang Du, Dikang Du, Yulun Du, Yu Fan, Yichen Feng, Kelin Fu, Bofei Gao, Hongcheng Gao, Peizhong Gao, Tong G… — Kimi K2: Open Agentic Intelligence
  51. M.-A-P. Team, Xinrun Du, Yifan Yao, Kaijing Ma, Bingli Wang, Tianyu Zheng, Kang Zhu, Minghao Liu, Yiming Liang, Xiaolong Jin, Zhenlin Wei, Chujie Zheng, Kaixin Deng, Shian Jia, Sichao Jiang, Yiyan Liao, Rui Li, Qinrui Li, Sirun Li, Yizhi Li, Yunwen Li, Dehua Ma, Yuansheng Ni, Haoran Que, Qiyao Wang,… find
  52. NLLB Team, Marta R. Costa-jussà, James Cross, Onur Çelebi, Maha Elbayad, Kenneth Heafield, Kevin Heffernan, Elahe Kalbassi, Janice Lam, Daniel Licht, Jean Maillard, Anna Sun, Skyler Wang, Guillaume Wenzek, Al Youngblood, Bapi Akula, Loic Barrault, Gabriel Mejia Gonzalez, Prangthip Hansanti, John H… find
  53. Qwen Team. Qwen3 technical report, 2025. URL https://arxiv.org/abs/2505.09388. ↗
  54. Jörg Tiedemann. Parallel data, tools and interfaces in opus. In Lrec, volume 2012, pp. 2214–2218, 2012. find
  55. Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, Łukasz Kaiser, and Illia Polosukhin. Attention is all you need. Advances in neural information processing systems, 30, 2017.
  56. Yubo Wang, Xueguang Ma, Ge Zhang, Yuansheng Ni, Abhranil Chandra, Shiguang Guo, Weiming Ren, Aaran Arulraj, Xuan He, Ziyan Jiang, Tianle Li, Max Ku, Kai Wang, Alex Zhuang, Rongqi Fan, Xiang Yue, and Wenhu Chen. Mmlu-pro: A more robust and challenging multi-task language understanding benchmark. In A…
  57. Jason Wei, Xuezhi Wang, Dale Schuurmans, Maarten Bosma, Brian Ichter, Fei Xia, Ed Chi, Quoc Le, and Denny Zhou. Chain-of-thought prompting elicits reasoning in large language models, 2023. URL https://arxiv.org/abs/2201.11903. ↗
  58. Yonghui Wu, Mike Schuster, Zhifeng Chen, Quoc V Le, Mohammad Norouzi, Wolfgang Macherey, Maxim Krikun, Yuan Cao, Qin Gao, Klaus Macherey, et al. Google’s neural machine translation system: Bridging the gap between human and machine translation. arXiv preprint arXiv:1609.08144, 2016. find
  59. Qiyuan Zhang, Fuyuan Lyu, Zexu Sun, Lei Wang, Weixu Zhang, Wenyue Hua, Haolun Wu, Zhihan Guo, Yufei Wang, Niklas Muennighoff, Irwin King, Xue Liu, and Chen Ma. A survey on test-time scaling in large language models: What, how, where, and how well?, 2025. URL https://arxiv.org/abs/2503.2 4235. ↗
  60. Wenhao Zhu, Hongyi Liu, Qingxiu Dong, Jingjing Xu, Shujian Huang, Lingpeng Kong, Jiajun Chen, and Lei Li. Multilingual machine translation with large language models: Empirical results and analysis. In Kevin Duh, Helena Gómez-Adorno, and Steven Bethard (eds.), Findings of the Association for Comput…

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/zheng-2025-hunyuan-mt-technical-report/