Cohere Says Command A+ Model Beats Google Translate, OpenAI's gpt-oss on WMT24++
Cohere said its Command A+ model delivered its strongest machine-translation results to date, outperforming Mistral Medium 3.5, DeepSeek, OpenAI's gpt-oss, Claude Opus 4.6 and Google Translate on WMT24++ scored with xCOMET-XL. The company cited leads of 2.4 points in French, 1.9 in Spanish and 0.9 in German, and said its Arabic score was 5 points ahead of Mistral Medium 3.5 and 10 points ahead of DeepSeek V4 Pro.
Cohere said the gains also extended to Korean, Japanese, Hebrew and Chinese, and argued that better translation quality should mean fewer corrections, stronger retrieval and more reliable multilingual agents. The company added that RWS still performs better on translation, while noting that it built that system with RWS.
From the sources (5 posts)
@cohereCommand A+ sets a new high for Cohere's machine translation capabilities. Opening a clear gap over open source peers Mistral Medium 3.5, DeepSeek, & OpenAI's gpt-oss, as well as Claude Opus 4.6. A+ also outperforms specialist systems like
@cohereThe improvements run wide. Across all major European languages, Command A+ consistently pulls ahead of competitors on WMT24++ (xCOMET-XL): 🇫🇷 +2.4 pts in French 🇪🇸 +1.9 pts in Spanish 🇩🇪 +0.9 pts in German Higher translation quality mean
@cohereThe story gets bigger beyond Europe. Command A+ makes major gains in high-impact non-Latin languages; outperforming Mistral Medium 3.5 in Korean, Japanese, Hebrew, Chinese, and Arabic. For Arabic, that gap becomes: +5 pts over Mistral Med
@nickfrosstCommand A+は、Cohereの機械翻訳能力において新たな高みを実現しました。 Il crée un écart marqué par rapport aux concurrents open source tels que Mistral Medium 3.5, DeepSeek et gpt-oss d'OpenAI, ainsi que Claude Opus 4.6. A+ también supera a sistemas especializados
@aidangomezRT @cohere: Command A+ sets a new high for Cohere's machine translation capabilities. Opening a clear gap over open source peers Mistral M…