The Qdrant MCP server provides semantic memory capabilities by integrating the Qdrant vector search engine with language models through the Model Context Protocol. It offers two main tools: `qdrant-store` for persisting information with optional metadata into Qdrant collections, and `qdrant-find` for retrieving semantically relevant information through vector search queries. The server solves the problem of enabling LLMs to maintain and access persistent, semantically-indexed memory across conversations by acting as a memory layer on top of a vector database.
claude mcp add --transport stdio qdrant-mcp-server-qdrant uvx mcp-server-qdrant