MCP Radar

Qdrant MCP Server

The Qdrant MCP Server lets your AI use a Qdrant vector store for "memory" and semantic retrieval — stored content is recalled by semantic similarity, not exact match.

For example, "store these technical notes and later find related ones by meaning" — the AI uses this server to write vectors and recall semantically. This is Qdrant's official server.

See this server's health data →

What it does for you

Connects your AI to Qdrant — semantic search and vector collection management, a memory backend for RAG apps.

Capabilities

  • Semantic similarity search
  • Store / query vectors

Once installed, try asking your AI

Say this directly in Claude / Cursor or any client with this server connected

  • Find the passages most relevant to 'refund policy' in the knowledge base

Install / Connect

# No npm package — install from source.
# See the repo's README for the exact command:
# https://github.com/qdrant/mcp-server-qdrant

Follow the official README; env and args vary per server.

Why use it

Good for: developers building RAG, giving the AI long-term memory, or needing semantic search.

Built by Qdrant officially; its health and maintenance data is shown below.

Similar / alternative servers

Frequently Asked Questions

Do I need to run Qdrant myself?

You need an accessible Qdrant instance (self-hosted or cloud) for the server to read/write.

How is it different from a normal database?

It retrieves by vector similarity — good for "find semantically similar content," not exact SQL queries.