Cohere

Build custom agents to automate complex workflows and business processes with AI-driven decision-making.

A family of multilingual research models covering 70+ languages, enabling AI applications for diverse global markets.

Frontier language models accessible via API for building and deploying AI applications. Includes capabilities for text generation, embeddings, and retrieval-augmented generation.

An enterprise-grade AI workspace platform designed for building custom agents to automate complex workflows, generate documents and reports, and access insights grounded in company data.

High-performance generative model with agentic capabilities, unified reasoning, tool orchestration, and multimodal intelligence. Supports 49 languages.

Cohere's fastest and most powerful language model, available as open-source. Designed for high-performance agentic, multimodal, and multilingual AI applications.

Frontier-grade language models optimized for business tasks including document generation, search, and reasoning.

Product offering referenced in navigation; specific details not provided on homepage.

Service to train Cohere models on proprietary data and create unique AI solutions tailored to specific use cases, needs, and infrastructure requirements.

AI-powered tool to quickly generate documents, reports, and business content from data and workflows.

Semantic text representation model that captures text meaning for document comparison, efficient retrieval through vectorization, and contextual pattern discovery.

Text embedding models for semantic search, similarity matching, and vector database integrations.

Turnkey AI platform for enterprises to build, customize, and deploy secure AI solutions with immediate impact.

Training and fine-tuning services for Cohere models on proprietary data to create unique, tailored AI solutions for specific use cases and infrastructure requirements.

Dedicated, Cohere-managed deployment option for secure model hosting within your infrastructure.

Sovereign AI workplace platform that integrates all tools on your terms with nothing outsourced. Provides unified control over AI deployment and management.

Cohere's first model designed specifically for developers, enabling code generation and development tasks.

Relevance-based ranking model that prioritizes most relevant documents, personalizes search results to user needs, and dynamically refines results based on interactions.

Semantic ranking models designed to improve search relevance and retrieval accuracy for RAG and search applications.

Flexible deployment models including VPC, on-premises, and Cohere-managed Model Vault to ensure data privacy and security while running AI models.

High-fidelity speech-to-text model that converts audio to accurate text. Supports 14 languages and integrates with generative and retrieval systems for end-to-end speech-driven workflows.

Arabic language support addition to Cohere's speech-to-text transcription model.