# Hy-MT2: A Family of Fast, Efficient and Powerful Multilingual Translation Models in the Wild

arXiv: https://arxiv.org/abs/2605.22064 · HTML: https://mlsanity.com/paper/2605.22064/

## Key facts

- Cited by 10 papers in the map (data through 14 Oct 2026).
- Builds on 9 of its 15 references that are in the map.
- First posted on arXiv 21 May 2026.

## Builds on (most cited first, 9 of 9)

- [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024)
- [No Language Left Behind: Scaling Human-Centered Machine Translation](https://mlsanity.com/paper/2207.04672/) (2022)
- [Instruction-Following Evaluation for Large Language Models](https://mlsanity.com/paper/2311.07911/) (2023)
- [Multi-IF: Benchmarking LLMs on Multi-Turn and Multilingual Instructions Following](https://mlsanity.com/paper/2410.15553/) (2024)
- [BitDistiller: Unleashing the Potential of Sub-4-Bit LLMs via Self-Distillation](https://mlsanity.com/paper/2402.10631/) (2024)
- [ParetoQ: Improving Scaling Laws in Extremely Low-bit LLM Quantization](https://mlsanity.com/paper/2502.02631/) (2025)
- [MaXIFE: Multilingual and Cross-lingual Instruction Following Evaluation](https://mlsanity.com/paper/2506.01776/) (2025)
- [AngelSlim: A more accessible, comprehensive, and efficient toolkit for large model compression](https://mlsanity.com/paper/2602.21233/) (2026)
- [Sherry: Hardware-Efficient 1.25-Bit Ternary Quantization via Fine-grained Sparsification](https://mlsanity.com/paper/2601.07892/) (2026)

## Built on by (most cited first, 10 of 10)

- [IFMTBench: A Comprehensive Benchmark for Multilingual Translation Instruction Following](https://mlsanity.com/paper/2605.28218/) (2026)
- [Cultivar: A Contrastive and Locale-Oriented Translation Benchmark for Investigating Contamination and Localisation Robustness](https://mlsanity.com/paper/2608.09766/) (2026)
- [Last Translation Benchmark](https://mlsanity.com/paper/2609.04173/) (2026)
- [HardMTBench: Stress-Testing Chinese-English Translation on Knowledge-Intensive Domains](https://mlsanity.com/paper/2605.28315/) (2026)
- [Studying quantization trade-offs for efficient inference deployment in machine translation](https://mlsanity.com/paper/2607.29397/) (2026)
- [Scalable Frequency- and Length-Aware Subdocument Deduplication for Large Language Model Pretraining](https://mlsanity.com/paper/2608.03089/) (2026)
- [Reference-Free Post-Training of Open Large Language Models for Multilingual Machine Translation](https://mlsanity.com/paper/2608.10812/) (2026)
- [Quit While You're Ahead: Quit for Efficient Candidate Generation in Machine Translation Reranking](https://mlsanity.com/paper/2609.00588/) (2026)
- [VISTA-Bench: Benchmarking Multilingual Image Translation with Image-Specific Rubrics](https://mlsanity.com/paper/2609.37287/) (2026)
- [Index-Translate: A Multilingual Translation Model Family -- Text, Speech, Controlled Dubbing, and Long-Document Translation](https://mlsanity.com/paper/2609.40181/) (2026)

## Abstract (arXiv)

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. 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 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. Moreover, when paired with AngelSlim's 1.25-bit extreme quantization for on-device deployment, the lightweight 1.8B model requires only 440 MB of storage and achieves a 1.5x inference speedup.