# mcp-server-qdrant

> MCP Server

This repository provides an example of a Model Context Protocol (MCP) server for Qdrant, a vector search engine, enabling seamless integration between LLM applications and the Qdrant database as a

## Overview

- **Category:** AI
- **Language:** Python
- **Stars:** 1505
- **Forks:** 42
- **Owner:** qdrant
- **GitHub:** https://github.com/qdrant/mcp-server-qdrant
- **Created:** 2024-12-02T12:07:39+00:00
- **Updated:** 2025-03-28T13:08:50+00:00
- **Source:** https://model-context-protocol.com/servers/qdrant-mcp-server-model-context-protocol

## Setup

## Setup

**Using uvx:**

1.  Run the server directly using `uvx` with environment variables for configuration.
2.  Specify the transport protocol using the `--transport` flag (options: `stdio`, `sse`).

```shell
QDRANT_URL="http://localhost:6333" \
COLLECTION_NAME="my-collection" \
EMBEDDING_MODEL="sentence-transformers/all-MiniLM-L6-v2" \
uvx mcp-server-qdrant
```

**Using Docker:**

1.  Build the Docker container: `docker build -t mcp-server-qdrant .`
2.  Run the container, mapping port 8000 and setting environment variables.

```bash
docker run -p 8000:8000 \
  -e QDRANT_URL="http://your-qdrant-server:6333" \
  -e QDRANT_API_KEY="your-api-key" \
  -e COLLECTION_NAME="your-collection" \
  mcp-server-qdrant
```

**Installing via Smithery:**

1.  Install using the Smithery CLI: `npx @smithery/cli install mcp-server-qdrant --client claude`

**Manual configuration of Claude Desktop:**

1.  Add a configuration block to the `claude_desktop_config.json` file, specifying the command, arguments, and environment variables for either remote or local Qdrant.

**Using with Cursor/Windsurf:**

1. Customize the tool descriptions using environment variables `TOOL_STORE_DESCRIPTION` and `TOOL_FIND_DESCRIPTION`.
2. Configure the MCP server in Cursor/Windsurf settings, pointing to the running server using the SSE transport protocol (e.g., `http://localhost:8000/sse`).
3. Consider creating Cursor rules to ensure MCP tools are used when the agent produces new code snippets.

## Tools

## Available Tools

1.  `qdrant-store` (Stores information in the Qdrant database with optional metadata.)
2.  `qdrant-find` (Retrieves relevant information from the Qdrant database based on a query.)