Cohere
@cohere
We worked with @RWSGroup to fine-tune our Command translation model, which improved language and cultural expertise. Now, that model is the “brain” that powers RWS’ Language Weaver AI translation solution.
3retweets11likes
View on XCohere and RWS Launch 100B+ Parameter Translation Model for Enterprise
Cohere· Updated
Cohere partnered with RWS Group to fine-tune a 100B+ parameter Command model that now powers the Language Weaver Pro translation platform. This specialized model outperforms general-purpose frontier models by combining massive scale with deep linguistic and cultural expertise.
Cohere, an enterprise AI model developer, collaborated with RWS Group, a global provider of language and content technology, to fine-tune a specialized version of the
Command model. This 100B+ parameter model powers Language Weaver Pro. The fine-tuning process (adapting a pre-trained model for specific tasks) focused on enhancing linguistic accuracy and cultural nuance.This release marks a shift toward massive, task-specific models that challenge general-purpose frontier systems. While general models handle broad tasks, this engine is benchmarked to outperform both generalists and specialized incumbents like DeepL. It solves the enterprise need for high-fidelity translation that preserves complex document formatting and legal structures.
You can access this capability through the Language Weaver Pro platform, which supports 32 languages. The system is designed for high-stakes enterprise workloads requiring data privacy and structural integrity. While a research version exists with open weights, the production-grade version is available via the RWS Group managed service.
Every HeadsUpAI update is written based on its original source and reviewed before it's published. Read our editorial standards →
