Tencent's open multilingual translation model family (1.8B, 7B, 30B-A3B MoE) covering 33 languages.
Updated EU status checked Last shown
Sources: the vendor's own pages (linked above) and creator videos at the linked timestamps. EU status is information, not legal advice; see how we check.
Hy-MT2 (HY-MT2) is a family of "fast-thinking" translation models from Tencent Hunyuan, released in three sizes: 1.8B, 7B and 30B-A3B (a mixture-of-experts model with about 3B active parameters). All sizes translate among 33 languages and follow translation instructions given in multiple languages, including formatting requirements.
Tencent reports that the 7B and 30B-A3B models outperform open models such as DeepSeek-V4-Pro and Kimi K2.6 on its translation evaluations in fast-thinking mode. It also reports that the 1.8B model surpasses mainstream commercial translation APIs, including those from Microsoft and Doubao, overall. These are vendor-reported results. The release also includes IFMTBench, a benchmark for translation instruction-following.
Weights are published on Hugging Face and ModelScope under Apache-2.0, with FP8 and GGUF variants. For on-device use, AngelSlim 1.25-bit quantization reduces the 1.8B model to about 440 MB. Tencent also provides a Hy-MT2-Translator Skill for agent integration. Tencent is partnering with WMT26 on its video subtitle translation task.
Watch the moment: the exact second top creators show it. Videos load only when you press play, via youtube-nocookie.com.
Family of open multilingual translation models (1.8B, 7B, 30B MoE with 3B active) supporting 33 languages and detailed instruction following for formatting. Outperforms larger open models on benchmarks; FP8 and GGUF versions released.
Closest neighbours by meaning, not just by category.
Model · Tencent
Tencent Hunyuan's real-time Chinese-English speech-to-speech simultaneous translation model with voice cloning.
Model · Tencent
Tencent's open-weights 295B Mixture-of-Experts model (21B active) for reasoning and agentic coding.
Model · Alibaba
Small 0.5B-parameter open-weights base language model from Alibaba's Qwen2.5 series, Apache 2.0 licensed.
Model · Tencent
Tencent's Hy4 multimodal model with visual understanding, computer use, 3D/Blender and knowledge-work abilities.
Model · OpenBMB
Dense 2B-parameter open-weights language model from OpenBMB for on-device and edge deployment, Apache-2.0 licensed.
Model
Compact 1.7B open-weights text LLM for on-device use, with adjustable thinking levels and English/Chinese support.
