Hunyuan-MT-7B - Tencent Mixed Meta Open Source Lightweight Translation Model

What is Hunyuan-MT-7B?

Hunyuan-MT-7B is a lightweight translation model launched by Tencent's Mixed Meta Team, with 7 billion parameters, supporting the mutual translation of 33 languages and 5 folk-Chinese languages/dialects, including Cantonese, Uyghur, and Tibetan. It took the 1st place in 30 out of 31 language competitions in the International Association for Computational Linguistics (ACL) WMT2025 competition, demonstrating excellent translation capabilities. The model can accurately understand online phrases, ancient poems, social conversations, etc., and translate meaningfully with context to provide more accurate translation results.

Functional Features of Hunyuan-MT-7B

• Multi-language support: Supports 33 languages and 5 folk-Chinese languages/dialects, covering a wide range of minor languages and dialects, to meet diversified translation needs.
• accurate translation: It can accurately understand linguistic expressions in special contexts such as online phrases, ancient poems, social conversations, etc., and provide more accurate translation results by combining Italian translation with the context.
• High performance: The number of references is only 7 billion, the inference speed is fast, and after compression the performance is further improved by 30%, which can handle more translation requests under the same hardware conditions.
• Deployment flexibility: It can operate in diverse hardware environments from high-end servers to edge devices, with low deployment costs for different application scenarios.
• wide range of applications: It has been accessed to a number of businesses such as Tencent Conference, Enterprise WeChat, etc., which helps to improve the product experience, and can also be applied to a number of fields such as education, foreign trade, culture and tourism, and scientific research.

Core Advantages of Hunyuan-MT-7B

• Translation excellence: Took 30 out of 31 1st place finishes in the International Association for Computational Linguistics (ACL) WMT2025 competition, demonstrating superior translation performance and the ability to accurately understand and translate multiple languages and dialects.
• Lightweight and efficient: Only 7 billion references, fast inference speed, processed by Tencent's self-developed AngelSlim compression tool, the performance is further enhanced by 30%, which is more capable of processing under the same hardware conditions.
• Extensive language support: Supports 33 languages and 5 ethnic Chinese languages/dialects, including Cantonese, Uyghur, Tibetan, etc., to meet the translation needs in a multi-language environment.
• Precise contextual understanding: It can accurately understand linguistic expressions in special contexts such as online phrases, ancient poems, social conversations, etc., and translate meaningfully with the context to provide more contextualized translation results.
• Low deployment costs: It can operate efficiently in diverse hardware environments, from high-end servers to edge devices, and can be well deployed with relatively lower deployment, operation and maintenance costs.

Hunyuan-MT-7B's official website address

• Official website address:: https://hunyuan.tencent.com/modelSquare/home/list
• Github:: https://github.com/Tencent-Hunyuan/Hunyuan-MT/
• HuggingFace:: https://huggingface.co/collections/tencent/hunyuan-mt-68b42f76d473f82798882597

People who use Hunyuan-MT-7B

• Students and educators: To help cross language barriers, access global educational resources, and enhance the efficiency of international academic exchange and learning.
• foreign trade practitioner: Facilitate international trade, help communicate effectively with global customers, and expand international markets.
• travelers: Facilitates the travel experience in different language environments, provides real-time translation support, and enhances cultural exchange.
• (scientific) researcher: Rapid access to and sharing of international cutting-edge research results, and promotion of transnational scientific cooperation and academic exchanges.
• Company employees: Supports cross-linguistic communication within enterprises, enhances the efficiency of internal collaboration within multinational enterprises, and facilitates globalized business development.
• Social platform users: Facilitates cross-language communication and improves communication and understanding between users from different language backgrounds.