MCP Server Qdrant Retrive

🚀 Qdrant Semantic Retrieval MCP Server

An MCP server for semantic search, paired with the Qdrant vector database.

🚀 Quick Start

This is an MCP server designed for semantic search, working in conjunction with the Qdrant vector database.

✨ Features

• Multi-collection semantic search
• Support for multiple queries
• Configurable number of results
• Collection source tracking

⚠️ Important Note

This server connects to a Qdrant instance specified by the URL.

⚠️ Important Note

The first retrieval may be slow because the MCP server will download the required embedding model.

📚 Documentation

API

Tools

• qdrant_retrieve
  • Retrieve semantically similar documents from multiple Qdrant vector storage collections based on multiple queries.
  • Input:
    • collectionNames (string[]): An array of Qdrant collection names to search.
    • topK (number): The number of most similar documents to retrieve (default: 3).
    • query (string[]): An array of query texts.
  • Return:
    • results: An array of retrieved documents, including:
      • query: The query that generated this result.
      • collectionName: The name of the collection where this result originated.
      • text: The content of the document body.
      • score: The similarity score (ranging from 0 to 1).

Usage with Claude Desktop

Add the following content to your claude_desktop_config.json:

{
  "mcpServers": {
    "qdrant": {
      "command": "npx",
      "args": ["-y", "@gergelyszerovay/mcp-server-qdrant-retrive"],
      "env": {
        "QDRANT_API_KEY": "your_api_key_here"
      }
    }
  }
}

Command-line Options

MCP server for semantic search with Qdrant vector database.

Options
  --enableHttpTransport      Enable HTTP transport [default: false]
  --enableStdioTransport     Enable standard input/output transport [default: true]
  --enableRestServer         Enable REST API server [default: false]
  --mcpHttpPort=<port>       Port number for the MCP HTTP server [default: 3001]
  --restHttpPort=<port>      Port number for the REST HTTP server [default: 3002]
  --qdrantUrl=<url>          URL of the Qdrant vector database [default: http://localhost:6333]
  --embeddingModelType=<type> Type of embedding model to use [default: Xenova/all-MiniLM-L6-v2]
  --help                     Display this help message

Environment Variables
  QDRANT_API_KEY            API key for authenticated Qdrant instances (optional)

Examples
  $ mcp-qdrant --enableHttpTransport
  $ mcp-qdrant --mcpHttpPort=3005 --restHttpPort=3006
  $ mcp-qdrant --qdrantUrl=http://qdrant.example.com:6333
  $ mcp-qdrant --embeddingModelType=Xenova/all-MiniLM-L6-v2