Command model family. Cohere’s flagship Command models (including Command R, Command R+, and related variants) deliver enterprise-grade generation, conversation, reasoning, and tool-use capabilities with long-context windows and support for retrieval-augmented generation (RAG), reranking, and structured outputs for downstream workflows.

Aya multilingual models. The Aya family provides broad multilingual coverage designed for global enterprises, enabling use cases such as multilingual customer support, content generation, and knowledge management across many languages rather than English alone.

Embed and rerank models. Cohere’s embedding and rerank models power semantic search, classification, clustering, and recommendation systems, allowing enterprises to build RAG pipelines and intelligent knowledge retrieval on their proprietary data.

RAG and knowledge-integration tooling. Native support for retrieval-augmented generation lets customers ground model outputs in internal documents, databases, and knowledge bases, reducing hallucinations and improving factuality in high-stakes settings.

Multimodal capabilities. Cohere is extending beyond text-only inputs into multimodal use cases, supporting text–image understanding so enterprises can process documents, diagrams, and other visual artifacts alongside language.

North workspace platform. North is Cohere’s secure enterprise AI workspace and agent platform for knowledge workers, moving up the stack from raw APIs to packaged interfaces that can be deployed inside organizations for search, summarization, and workflow automation.

Private and sovereign deployment. Cohere supports on-premises deployments, sovereign cloud environments, and cloud-agnostic configurations, including options where models run inside a customer’s own VPC or air-gapped infrastructure.

Enterprise APIs and tooling. A comprehensive API surface for generation, summarization, classification, embeddings, and reranking is designed to plug into existing enterprise infrastructure, including CRM, document management, analytics, and custom internal applications.