Qdrant MCP Server
An MCP server that gives AI assistants semantic search over a local Qdrant vector database, with AST-aware codebase indexing, git history search, hybrid keyword and vector search, and support for several embedding providers.
This MCP server connects an MCP client to a Qdrant vector database for semantic search, working out of the box with Ollama for local, privacy-preserving embeddings so no data leaves the machine, or with OpenAI, Cohere or Voyage AI when a hosted embedding provider is preferred. Beyond plain document search, it offers intelligent codebase indexing with AST-aware chunking and semantic code search, indexing of git commit history for searching past changes and fixes, and contextual search that correlates code and git history together, as well as federated search across multiple indexed repositories.
Other features include incremental indexing that only re-embeds changed files, hybrid search combining semantic and keyword matching, configurable prompts for guided workflows, rate limiting with exponential backoff, structured JSON logging, and either a local stdio transport or a remote HTTP deployment with API-key authentication for secured or cloud-hosted Qdrant instances. It requires Node.js 22 or 24 and Podman or Docker with Compose support, and is written in TypeScript.