Cohere tops WMT26 translation benchmark: small agentic model scores 84.36, beating Mistral and DeepL
cohere · x · 2026-10-15
Cohere shared WMT26 all-languages translation benchmark results: North Small Translate (Agentic) leads with 84.36, followed by its non-agentic version at 83.60, ahead of Mistral Large 3 (81.57), Qwen 3.5 397B (81.56) and DeepL NextGen (81.37). Google Translate trails at 68.20, below Command A Translate (69.95), with Nemotron 3 Ultra (67.08) and Qwen 3.6 27B (61.93) at the bottom — a sign that small specialized translation models now clearly beat general-purpose LLMs and dedicated translation services.
More from Models
- Cloudflare launches Clef-omni: one pipeline for audio, video, image and text decisions — michellechen · 2026-10-15
- NVIDIA teases its own model launch amid a week packed with releases — TheOyinbooke · 2026-10-15
- Silence Is Billable on Voice Models: Even Dead Air Costs Money — sophiamyang · 2026-10-15
- NCM: new text generation paradigm up to 31x faster than diffusion LMs with better quality — simonguozirui · 2026-10-15
- Gemma 4 26B priced at $0.07/M input tokens, ~40% below comparable providers — testingcatalog · 2026-10-15
- Epoch AI: GPT-6 Astra tops frontier models in Slay the Spire 2 but still loses to top humans — sandersted · 2026-10-15