腾讯混元开源 Hy-MT2 多语言翻译模型系列
腾讯混元在 HuggingFace 与 ModelScope 开源 Hy-MT2 翻译模型系列，包含 1.8B、7B 和 30B-A3B（MoE）三个规模，支持 33 种语言，并强调多语言翻译指令跟随能力。团队称 7B 与 30B-A3B 在“快思考”模式下优于 DeepSeek-V4-Pro、Kimi K2.6 等开源模型，1.8B 版本整体超过 Microsoft、Doubao 等商业 API，这些结论均来自官方评测，尚未经第三方验证。发布同时开源 IFMTBench 指令跟随评测集与 Hy-MT2-Translator Skill，并借助 AngelSlim 1.25-bit 量化把 1.8B 模型压缩到约 440MB、推理速度提升 1.5 倍。腾讯混元还宣布与 WMT26 合作，为使用 Hy-MT 系列参赛的队伍设立专项奖项。
把翻译模型压到端侧可部署的体量并开放权重，可能让本地化和多语种翻译场景更多转向自托管方案。
English ｜ 中文
🖥️ Official Website | 💬 GitHub | 🪡 AngelSlim | 📚 Hy-MT2 Report
Model Introduction
Hy-MT2 is a family of “fast-thinking” multilingual translation models designed for complex real-world scenarios. It includes three model sizes: 1.8B, 7B, and 30B-A3B (MoE), all of which support translation among 33 languages and effectively follow translation instructions in multiple languages. For on-device deployment, AngelSlim 1.25-bit extreme quantization reduces the storage requirement of the 1.8B model to only 440 MB and improves inference speed by 1.5x. Multi-dimensional evaluations show that Hy-MT2 delivers outstanding performance across general, real-world business, domain-specific, and instruction-following translation tasks. The 7B and 30B-A3B models outperform open-source models such as DeepSeek-V4-Pro and Kimi K2.6 in fast-thinking mode, while the lightweight 1.8B model also surpasses mainstream commercial APIs from providers such as Microsoft and Doubao overall.
#Tencent Hunyuan#Hy-MT2#AngelSlim#IFMTBench#WMT26#Hugging Face#ModelScope