The Cohere MCP server is an implementation of the Model Context Protocol that connects your AI agent and assistants like Claude, Cursor, etc directly to your Cohere account. It provides structured and secure access to Cohere chat, embedding, reranking, transcription, and dataset workflows, so your agent can generate chat responses, create embeddings, rank documents, transcribe audio, and manage processing jobs on your behalf.
- Chat and structured responses: Have your agent generate responses from ordered conversations, use supporting documents, call tools, apply reasoning controls, or return structured output.
- Embeddings and token processing: Direct your agent to create embeddings from text, images, or structured multimodal content, and tokenize or detokenize text with a selected Cohere model.
- Document relevance ranking: Let the agent rerank text documents against a query and return the most relevant source content.
- Audio transcription: Instruct your agent to transcribe uploaded audio files into text using a Cohere transcription model.
- Datasets and processing jobs: Have your agent upload and inspect datasets, create or cancel batch and embedding jobs, check their progress, and review available Cohere models.