Milvus Integration

v1.0.0Databasesstable

An integration server implementing the Model Context Protocol that enables LLM applications to interact with Milvus vector database functionality, allowing vector search, collection management, and data operations through natural language.

mcp-server-for-milvusmcpai-integration
Share:
231
Stars
0
Downloads
0
Weekly
0/5

What is Milvus Integration?

Milvus Integration is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to integration server implementing the model context protocol that enables llm applications to interact with milvus vector database functionality, allowing vector search, collection management, and data ...

An integration server implementing the Model Context Protocol that enables LLM applications to interact with Milvus vector database functionality, allowing vector search, collection management, and data operations through natural language.

This server falls under the Databases category on MCPgee, the world's largest MCP server directory with 33,000+ servers.

Features

  • An integration server implementing the Model Context Protoco

Use Cases

Execute vector search and collection management operations.
Store and retrieve embeddings from Milvus.
Enable semantic search for LLM applications.
zilliztech

Maintainer

LicenseApache 2.0
Languagepython
Versionv1.0.0
UpdatedMay 21, 2026
Statushealthy
Maintenanceactive

Works with

ClaudeOpenAIwindowsmacoslinux

Installation

Manual Installation

npx mcp-server-for-milvus

Configuration

Configuration Details

Config File

claude_desktop_config.json

Performance

Response Metrics

Response Time< 200ms
ThroughputMedium

Resource Usage

Memory UsageLow
CPU UsageLow

How to Set Up and Use Milvus Integration

MCP Server for Milvus is an integration layer that allows LLM applications to interact with a Milvus vector database through the Model Context Protocol, enabling natural language-driven vector search, collection management, data insertion, and deletion without writing client code. It connects any MCP-compatible AI client such as Claude Desktop or Cursor directly to a local or remote Milvus instance, making it straightforward to build semantic search, RAG (retrieval-augmented generation), and embedding management workflows. The server supports stdio, SSE, and streamable HTTP transport modes to fit a range of deployment architectures.

Prerequisites

  • Python 3.10 or higher
  • uv package manager installed (https://github.com/astral-sh/uv)
  • A running Milvus instance (local via Docker or remote; default port 19530)
  • The mcp-server-milvus repository cloned locally
  • An MCP-compatible client such as Claude Desktop or Cursor
1

Start a Milvus instance

If you do not already have Milvus running, the quickest way is to start it with Docker Compose. The default URI is http://localhost:19530.

# Using Milvus standalone with Docker
curl -sfL https://raw.githubusercontent.com/milvus-io/milvus/master/scripts/standalone_embed.sh -o standalone_embed.sh
bash standalone_embed.sh start
2

Clone the MCP server repository

Clone the server source code to a local directory. The uv-based runner does not require a separate pip install.

git clone https://github.com/zilliztech/mcp-server-milvus.git
cd mcp-server-milvus
3

Test the server in stdio mode

Run the server directly to confirm it connects to Milvus before configuring your MCP client. Replace the URI if your Milvus is on a different host or port.

uv run src/mcp_server_milvus/server.py --milvus-uri http://localhost:19530
4

Configure Claude Desktop for stdio mode

Edit the Claude Desktop config file at ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) and add the Milvus server entry. Update the directory path to where you cloned the repo and the milvus-uri to match your instance.

{
  "mcpServers": {
    "milvus": {
      "command": "uv",
      "args": [
        "--directory",
        "/path/to/mcp-server-milvus/src/mcp_server_milvus",
        "run",
        "server.py",
        "--milvus-uri",
        "http://localhost:19530"
      ]
    }
  }
}
5

Restart Claude Desktop

Quit and reopen Claude Desktop so it picks up the new MCP server configuration. The Milvus tools should now be available in the tool list.

Milvus Integration Examples

Client configuration

Claude Desktop stdio mode configuration. Replace /path/to/mcp-server-milvus with the actual clone directory and adjust the milvus-uri if your instance is remote.

{
  "mcpServers": {
    "milvus": {
      "command": "uv",
      "args": [
        "--directory",
        "/path/to/mcp-server-milvus/src/mcp_server_milvus",
        "run",
        "server.py",
        "--milvus-uri",
        "http://localhost:19530"
      ]
    }
  }
}

Prompts to try

Natural language prompts to use with Claude Desktop once the Milvus MCP server is connected.

- "List all collections in my Milvus database."
- "Create a new collection called 'product_docs' with a 768-dimensional float vector field."
- "Search the 'knowledge_base' collection for documents semantically similar to 'machine learning optimization techniques', return the top 5 results."
- "Insert these three text embeddings into the 'articles' collection."
- "Show me the schema and entity count for the 'customer_feedback' collection."
- "Delete all entities in the 'temp_vectors' collection where the source field equals 'batch_2024_01'."

Troubleshooting Milvus Integration

Server fails to start with 'Connection refused' or 'Failed to connect to Milvus'

Verify that your Milvus instance is running and accessible at the URI you provided. Run 'curl http://localhost:19530/v1/vector/collections' to test connectivity. If using Docker, ensure the container is up with 'docker ps' and that port 19530 is exposed.

The .env file settings are being ignored in favor of command-line arguments

This is by design — the .env file in src/mcp_server_milvus/ takes higher priority than command-line arguments. Edit the .env file to set MILVUS_URI and any authentication tokens instead of passing them on the command line.

Claude Desktop does not show Milvus tools after restarting

Verify the uv binary path is correct by running 'which uv' in your terminal. The command in the JSON config must be the full path or a name resolvable from Claude Desktop's environment. Also double-check the --directory path points to the src/mcp_server_milvus subdirectory, not the repository root.

Frequently Asked Questions about Milvus Integration

What is Milvus Integration?

Milvus Integration is a Model Context Protocol (MCP) server that integration server implementing the model context protocol that enables llm applications to interact with milvus vector database functionality, allowing vector search, collection management, and data operations through natural language. It connects AI assistants to external tools and data sources through a standardized interface.

How do I install Milvus Integration?

Follow the installation instructions on the Milvus Integration GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.

Which AI clients work with Milvus Integration?

Milvus Integration works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.

Is Milvus Integration free to use?

Yes, Milvus Integration is open source and available under the Apache 2.0 license. You can use it freely in both personal and commercial projects.

Browse More Databases MCP Servers

Explore all databases servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.

Quick Config Preview

{ "mcpServers": { "mcp-server-for-milvus": { "command": "npx", "args": ["-y", "mcp-server-for-milvus"] } } }

Add this to your claude_desktop_config.json or .cursor/mcp.json

Read the full setup guide →

Ready to use Milvus Integration?

Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.

33,000+ ServersFree & Open SourceStep-by-Step Guides