We introduce Omnilingual MT, a family of multilingual translation (MT) systems covering more than 1,600 languages—the broadest benchmarked coverage of any MT system to date. We make contributions in several key research questions in the field: strategies for the use of monolingual data, curating seed training data, tokenization and suitable architectures, the competitiveness of small specialized MT models as opposed to general purpose large models, and evaluation protocols. Our data strategy combines large-scale multilingual corpora with backtranslation for 2,000+ languages and MeDLEy, a grammatically-diverse seed dataset for 109 low-resource languages. We explore two complementary architectures: OMT-LLAMA, a decoder-only model (1B/3B/8B) built on LLAMA3 with multilingual continual pretraining, and OMT-NLLB, a 3B encoder-decoder model with a novel three-stage training strategy exploiting non-parallel data via autoencoding. For evaluation, we extend the BOUQuET benchmark to 275 languages, develop Met-BOUQuET with human annotations of machine quality in 161 directions, and BLASER 3, a reference-free quality estimation metric. Our specialized 1B–8B models match or exceed 70B general-purpose LLMs, demonstrating that specialization is a more reliable path to multilingual translation than scale alone. Omnilingual MT achieves “passable” translation quality as judged by human annotators for over 440 languages, doubling it over the previous state of the field at ∼200 languages, and translates in zero-shot mode into ∼1,200 languages versus 300–400 for the baselines. With Omnilingual MT, we showcase a path encompassing the full pipeline of data and system development towards designing models that are multilingual from the onset.

We propose Mixture of Languages (MoL), a new strategy to pretrain largely multilingual encoders. Recent work in this field has relied on training transformer encoders on a large amount of multilingual data, with all parameters shared across all languages, without studying how to optimally balance language transfer and interference to achieve better performance. To address this, MoL proposes to group languages based on their similarity, and add parallel, sparsely activated layers that process each group independently. This architecture allows MoL to boost language transfer while minimizing interference, without increasing the active parameter count. We show that MoL largely outperforms a dense counterpart trained with the same configuration, as well as MoE models and public multilingual encoders such as XLM-R or mBERT on downstream tasks.

Cross-lingual sentence encoders (CLSE) create fixed-size sentence representations with aligned translations. Current pre-trained CLSE approaches use sentence-level objectives only. This can lead to loss of information, especially for tokens, which then degrades the sentence representation. We propose MEXMA, a novel approach that integrates both sentence-level and token-level objectives. The sentence representation in one language is used to predict masked tokens in another language, with both the sentence representation and all tokens directly update the encoder. We show that adding token-level objectives greatly improves the sentence representation quality across several tasks. Our approach outperforms current pre-trained cross-lingual sentence encoders on bitext mining as well as several downstream tasks. We also analyse the information encoded in our tokens, and how the sentence representation is built from them.