Cohere's open-weight 218B translation model beats DeepL on WMT26.

North Small Translate is a mixture-of-experts model with 218B total and 25B active parameters, covering 50 languages. It scores 83.60 on WMT26, and an agentic multi-pass variant scores 84.36. DeepL NextGen scores 81.37 and Google Translate 68.20. On book-length translation in a single call it scores 48.9 to Google's 21.3. The weights are CC BY-NC 4.0, so commercial use goes through Cohere or RWS. YuE2-3B, a new open music model whose best-of-8 score edges past Suno v5, has the same non-commercial license.