Qdrant MCP Server
✓ Text source-checked Oct 31, 2026Official Qdrant MCP server that acts as a semantic memory layer for keeping and retrieving memories in the Qdrant vector search engine.
Source-checked against the vendor's site · rebuilt from the vendor's own pagesProfile updated Link checked How we rate
In short
Qdrant MCP Server is a free, official semantic memory layer for Qdrant. It enables storing and retrieving memories via FastEmbed models for MCP clients like Cursor and Claude.
The Qdrant MCP Server is an official Model Context Protocol implementation that serves as a semantic memory layer on top of the Qdrant vector search engine. The server is compatible with any MCP client, including Cursor, VS Code, Claude Desktop, and Windsurf.
The qdrant-store tool allows users to store information and optional metadata into a specified Qdrant collection. The qdrant-find tool retrieves relevant information from the Qdrant database based on a search query.
The vendor says that the server uses the sentence-transformers/all-MiniLM-L6-v2 embedding model by default and currently supports only FastEmbed models. Configuration is managed through environment variables, with the --transport flag as the sole command-line argument for selecting the protocol. The server supports stdio, SSE, and streamable-http transport protocols, with stdio as the default for local clients and SSE or streamable-http for remote connections.
Users can run the server directly using uvx without specific installation steps. A Dockerfile is provided for building and running the MCP server in a containerized environment. The server supports a local Qdrant database path as an alternative to a remote Qdrant URL.
The project is licensed under the Apache License 2.0.
According to Qdrant MCP Server: The Qdrant MCP Server is an official Model Context Protocol implementation that serves as a semantic memory layer on top of the Qdrant vector search engine. Its pages list this strength: Official implementation from the Qdrant team. They also state this limitation: Only FastEmbed models are currently supported for embeddings.
— The AI Tool Bible editorial team
Pros
- ✅ Official implementation from the Qdrant team.
- ✅ Compatible with any MCP-compatible client.
- ✅ Supports local Qdrant database paths for offline or local development.
- ✅ Includes a read-only mode that disables the store tool.
Cons
- ⚠️ Only FastEmbed models are currently supported for embeddings.
- ⚠️ Users cannot provide both a remote Qdrant URL and a local path simultaneously.
Facts
- API available
- Yes (not source-checked)
- Open source
- Yes (license: Apache-2.0)
- GitHub stars
- 1,559 (checked Oct 29, 2026) (not source-checked)
- Last GitHub push
- Sep 4, 2026 (not source-checked)
- First commit
- Dec 2024 (not source-checked)
- Model used
- FastEmbed (default: sentence-transformers/all-MiniLM-L6-v2); pairs with any MCP-capable LLM such as Claude 3.5/4, GPT-4o, or local models (not source-checked)
Where this tool sits
Use cases
Frequently asked
- How much does the Qdrant MCP Server cost?
- The Qdrant MCP Server is free. The provided documentation does not list any specific pricing details or subscription fees for using the server or its associated tools.
- Which LLMs and embedding models are supported?
- It pairs with any MCP-capable LLM, such as Claude 3.5/4, GPT-4o, or local models. For embeddings, it uses FastEmbed by default, specifically the sentence-transformers/all-MiniLM-L6-v2 model.
- How do I install and run the server?
- You can run the server directly using uvx without specific installation steps. A Dockerfile is also provided for building and running the MCP server in a containerized environment if preferred.
- What transport protocols does it support?
- The server supports stdio, SSE, and streamable-http transport protocols. stdio is the default for local clients, while SSE or streamable-http are used for remote connections. The --transport flag selects the protocol.
- Can I use it with local Qdrant databases?
- Yes, the server supports a local Qdrant database path as an alternative to a remote Qdrant URL. Configuration is managed through environment variables to specify the connection details.