FrontierSeptember 11, 2026via MarkTechPost
Cohere Releases North Small Translate: A 218B MoE Translation Model That Scores 83.6 on WMT26 Across 50 Languages
Why it matters
A specialized frontier model release demonstrating MoE efficiency in a narrow domain (translation). Practitioners evaluating multilingual capabilities and open-weight alternatives gain a concrete benchmark; enthusiasts track Cohere's competitive positioning against general-purpose frontier labs in capability-focused releases.
Key signals
- Model: North Small Translate, 218B parameters total, 25B active per token (MoE architecture)
- Benchmark: 83.6 on WMT26 evaluation
- Language coverage: 50 languages
- Licensing: open-weight (non-commercial free), commercial via Cohere Model Vault or RWS Language Weaver
- Architecture: Mixture-of-Experts (MoE) specialization for translation task
The hook
Cohere ships a 218B MoE translation model scoring 83.6 on WMT26 across 50 languages — open weights for non-commercial use.
Cohere has released North Small Translate, an open-weight Mixture-of-Experts model built for machine translation across 50 languages. It uses 25B of its 218B parameters per token and scores 83.6 on Cohere's WMT26 evaluation. Weights are free for non-commercial use, with commercial access through Coh…