Skip to main content
Watch: I Built an AI App Without Leaving Databricks

Databases - official Qdrant server

Qdrant MCP server

Official Qdrant MCP server that acts as a semantic memory layer: it stores information with optional metadata in a Qdrant collection and finds it again by semantic search, using FastEmbed models for embeddings. It runs locally from mcp-server-qdrant (pypi) with API key auth.

Last verified 2 tools8 sources

Install

Claude Code

About Claude Code
Terminal
claude mcp add code-search \
-e QDRANT_URL="http://localhost:6333" \
-e COLLECTION_NAME="code-repository" \
-e EMBEDDING_MODEL="sentence-transformers/all-MiniLM-L6-v2" \
-e TOOL_STORE_DESCRIPTION="Store code snippets with descriptions. The 'information' parameter should contain a natural language description of what the code does, while the actual code should be included in the 'metadata' parameter as a 'code' property." \
-e TOOL_FIND_DESCRIPTION="Search for relevant code snippets using natural language. The 'query' parameter should describe the functionality you're looking for." \
-- uvx mcp-server-qdrant

VS Code

JSON
{
  "inputs": [
    {
      "type": "promptString",
      "id": "qdrantUrl",
      "description": "Qdrant URL"
    },
    {
      "type": "promptString",
      "id": "qdrantApiKey",
      "description": "Qdrant API Key",
      "password": true
    },
    {
      "type": "promptString",
      "id": "collectionName",
      "description": "Collection Name"
    }
  ],
  "servers": {
    "qdrant": {
      "command": "uvx",
      "args": ["mcp-server-qdrant"],
      "env": {
        "QDRANT_URL": "${input:qdrantUrl}",
        "QDRANT_API_KEY": "${input:qdrantApiKey}",
        "COLLECTION_NAME": "${input:collectionName}"
      }
    }
  }
}

Generic

JSON
{
  "qdrant": {
    "command": "uvx",
    "args": ["mcp-server-qdrant"],
    "env": {
      "QDRANT_URL": "https://xyz-example.eu-central.aws.cloud.qdrant.io:6333",
      "QDRANT_API_KEY": "your_api_key",
      "COLLECTION_NAME": "your-collection-name",
      "EMBEDDING_MODEL": "sentence-transformers/all-MiniLM-L6-v2"
    }
  }
}

Tools

qdrant-store
Store information, with optional metadata, in a Qdrant collection.
qdrant-find
Retrieve relevant information from the Qdrant database by semantic query.

As documented on October 17, 2026. A running server reports its current list through tools/list.

Notes and caveats

  • Self-hosted only: Qdrant runs no hosted endpoint.
  • Configured through environment variables.
  • Works with a remote Qdrant (QDRANT_URL plus an optional QDRANT_API_KEY) or a local path (QDRANT_LOCAL_PATH), and creates the collection if it does not exist.
  • Embeddings are FastEmbed models only.
  • QDRANT_READ_ONLY=true turns off qdrant-store.
  • The Dockerfile is build-it-yourself, with no published image.
  • The Claude Code snippet in the docs sets the server up for code search.
  • For Cursor the docs only describe pointing at a running SSE server (http://localhost:8000/sse) and give no config block.
  • Latest release v0.8.1 (2025-12-10).
  • The README calls it 'official' but gives no beta/GA label.

More MCP servers for databases

Related reading

FAQ

What does the Qdrant MCP server do?

Official Qdrant MCP server that acts as a semantic memory layer: it stores information with optional metadata in a Qdrant collection and finds it again by semantic search, using FastEmbed models for embeddings. The vendor documents 2 tools, including qdrant-store and qdrant-find.

How do I add the Qdrant MCP server to Claude Code?

Qdrant documents a Claude Code config; copy it from the Claude Code section of this page. Snippets for VS Code and Generic are there too.

Does the Qdrant MCP server need authentication?

Yes, API key. Transport: stdio, SSE, Streamable HTTP. Status: Unlabelled, last checked October 17, 2026.

Sources

Last verified against 8 Qdrant pages, then re-checked independently.

Newsletter

New MCP servers, the week they ship

One email per week. New official MCP servers, the coding tools that support them and how developers wire them up.