# Omnilingual MT: Machine Translation for 1,600 Languages

arXiv: https://arxiv.org/abs/2603.16309 · HTML: https://mlsanity.com/paper/2603.16309/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 17 Mar 2026.

## Built on by (most cited first, 5 of 5)

- [ALEE: Any-Language Evaluation of Embeddings via English-Centric Minimal Pairs](https://mlsanity.com/paper/2607.00171/) (2026)
- [Evaluating In-Context Translation with Synchronous Context-Free Grammar Transduction](https://mlsanity.com/paper/2604.07320/) (2026)
- [There is No Theoretical Curse of Multilinguality For Embedding Space Structure](https://mlsanity.com/paper/2608.17088/) (2026)
- [Improving Cross-Lingual Token Representations by Adding a Pinch of SALT](https://mlsanity.com/paper/2609.09953/) (2026)
- [The Constitution Speaks Nheengatu: An Open MT System and Reproducible Corpus Pipeline for the Amazonian L'ingua Geral](https://mlsanity.com/paper/2610.14760/) (2026)

## Abstract (arXiv)

High-quality machine translation (MT) can scale to hundreds of languages, setting a high bar for multilingual systems. However, compared to the world's 7,000 languages, current systems still offer only limited coverage: about 200 languages on the target side, and maybe a few hundreds more on the source side, supported due to cross-lingual transfer. And even these numbers have been hard to evaluate due to the lack of reliable benchmarks and metrics. We present Omnilingual Machine Translation (OMT), the first MT system supporting more than 1,600 languages. This scale is enabled by a comprehensive data strategy that integrates large public multilingual corpora with newly created datasets, including manually curated MeDLEY bitext. We explore two ways of specializing a Large Language model (LLM) for machine translation: as a decoder-only model (OMT-LLaMA) or as a module in an encoder-decoder architecture (OMT-NLLB). Notably, all our 1B to 8B parameter models match or exceed the MT performance of a 70B LLM baseline, revealing a clear specialization advantage and enabling strong translation quality in low-compute settings. Moreover, our evaluation of English-to-1,600 translations further shows that while baseline models can interpret undersupported languages, they frequently fail to generate them with meaningful fidelity; OMT-LLaMA models substantially expand the set of languages for which coherent generation is feasible. Additionally, OMT models improve in cross-lingual transfer, being close to solving the "understanding" part of the puzzle in MT for the 1,600 evaluated. Our leaderboard and main human-created evaluation datasets (BOUQuET and Met-BOUQuET) are dynamically evolving towards Omnilinguality and freely available.