MCP Server for Qdrant

What is MCP Server for Qdrant?

MCP Server for Qdrant is a Machine Control Protocol (MCP) server that stores and retrieves text information from a Qdrant vector database. It enables semantic search by generating embeddings via FastEmbed and supports optional metadata attachment. This server is intended for developers building AI‑powered applications that need persistent, searchable memory backed by a vector database.

How to use MCP Server for Qdrant?

Install with pip install mcp-server-qdrant or clone the repository and run make setup. Configure required environment variables (QDRANT_URL, QDRANT_API_KEY, COLLECTION_NAME) in a .env file. Start the server with python -m mcp_server_qdrant.main or make run; alternatively use docker-compose up. Two MCP tools are exposed: qdrant-store (store text with optional metadata) and qdrant-find (semantic search by query).

Key features of MCP Server for Qdrant

• Store text information with optional JSON metadata
• Semantic search over stored information
• FastEmbed integration for text embeddings
• Environment-based configuration via .env file
• Docker support for containerized deployment
• Exposes MCP tools: qdrant-store and qdrant-find

Use cases of MCP Server for Qdrant

• Provide persistent memory for AI agents that need to recall past conversations or data
• Build a semantic knowledge base that can be queried by natural language
• Enable retrieval‑augmented generation (RAG) pipelines with a Qdrant backend
• Store and search notes or documents with metadata filters
• Prototype MCP‑based applications that require vector storage

FAQ from MCP Server for Qdrant

What is MCP Server for Qdrant and how is it different from using Qdrant directly?

It provides an MCP interface for Qdrant, allowing any MCP‑compatible client to store and search vectors without writing custom integration code. The server handles embedding generation and tool definition.

How do I install MCP Server for Qdrant?

Install via pip (pip install mcp-server-qdrant) or clone the repository and run make setup from the source directory.

What dependencies or runtime requirements are needed?

You need a running Qdrant instance and the required environment variables (QDRANT_URL, QDRANT_API_KEY, COLLECTION_NAME). The server uses FastEmbed for embeddings; the model can be configured via EMBEDDING_MODEL.

Where is data stored?

Data is stored in the Qdrant database configured through QDRANT_URL and QDRANT_API_KEY, inside the specified COLLECTION_NAME. No local file storage is used by the server itself.

Does MCP Server for Qdrant support Docker?

Yes. A docker-compose.yml is provided; run docker-compose up to start the server alongside a Qdrant instance if configured.