Zilliz MCP

v1.0.0Databasesstable

Enables AI agents to interact with Milvus vector databases and Zilliz Cloud through natural language, allowing users to create clusters, manage collections, insert vector data, and perform semantic searches directly from their AI assistants.

zilliz-mcp-servermcpai-integration
Share:
34
Stars
0
Downloads
0
Weekly
0/5

What is Zilliz MCP?

Zilliz MCP is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to enables ai agents to interact with milvus vector databases and zilliz cloud through natural language, allowing users to create clusters, manage collections, insert vector data, and perform semantic se...

Enables AI agents to interact with Milvus vector databases and Zilliz Cloud through natural language, allowing users to create clusters, manage collections, insert vector data, and perform semantic searches directly from their AI assistants.

This server falls under the Databases and Data Science & ML categories on MCPgee, the world's largest MCP server directory with 33,000+ servers.

Features

  • Enables AI agents to interact with Milvus vector databases a

Use Cases

Query Milvus vector databases using natural language.
Manage Zilliz Cloud clusters and collections.
Perform semantic searches and store embeddings.
zilliztech

Maintainer

LicenseApache 2.0
Languagepython
Versionv1.0.0
UpdatedMay 17, 2026
Statushealthy
Maintenanceactive

Works with

ClaudeOpenAIwindowsmacoslinux

Installation

Manual Installation

npx zilliz-mcp-server

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 Zilliz MCP

The Zilliz MCP Server gives AI assistants natural-language control over both Zilliz Cloud (the managed Milvus service) and self-hosted Milvus vector databases. Through its tools, agents can manage the full lifecycle of a vector database deployment: create and suspend cloud clusters, define collections with custom schemas, insert and delete vector entities, and run semantic, scalar, and hybrid searches. Data engineers and AI developers building retrieval-augmented generation (RAG) pipelines can use this server to interact with their vector stores conversationally without writing SDK code.

Prerequisites

  • Python 3.10 or later and uv or uvx (pip install uv)
  • A Zilliz Cloud account with an API token (ZILLIZ_CLOUD_TOKEN), or a self-hosted Milvus instance
  • Internet access to reach api.zilliz.com for cloud operations
  • An MCP client such as Claude Desktop
1

Obtain a Zilliz Cloud API token

Log in to cloud.zilliz.com, navigate to your organization settings, and generate an API key. Copy the token — it will be used as ZILLIZ_CLOUD_TOKEN.

2

Install uvx if not already available

uvx is a tool runner bundled with uv that can run the server without a manual install step.

pip install uv
3

Test the server from the command line

Run the server once to verify it starts and can authenticate. You should see it initialize without errors.

ZILLIZ_CLOUD_TOKEN=your-token-here uvx zilliz-mcp-server
4

Add the server to Claude Desktop

Edit claude_desktop_config.json (macOS: ~/Library/Application Support/Claude/claude_desktop_config.json) and add the mcpServers block shown below.

5

Restart Claude Desktop and verify

After restarting, ask Claude to list your Zilliz Cloud projects or clusters to confirm the server is connected.

Zilliz MCP Examples

Client configuration

Add to claude_desktop_config.json. ZILLIZ_CLOUD_TOKEN is the only required credential for cloud usage.

{
  "mcpServers": {
    "zilliz-mcp-server": {
      "command": "uvx",
      "args": ["zilliz-mcp-server"],
      "env": {
        "ZILLIZ_CLOUD_TOKEN": "your-zilliz-cloud-token-here"
      }
    }
  }
}

Prompts to try

Operations covering cluster management, collection CRUD, and vector search.

- "List all my Zilliz Cloud projects and clusters"
- "Create a free Zilliz Cloud cluster named 'rag-dev' in the AWS us-east-1 region"
- "Create a collection called 'documents' with a 1536-dimensional float vector field and a text metadata field"
- "Insert 10 sample vectors into the documents collection"
- "Run a semantic search in the documents collection for vectors similar to my query embedding and return the top 5 results"
- "Suspend my dev cluster to save costs"

Troubleshooting Zilliz MCP

Authentication error or 'invalid token' when connecting to Zilliz Cloud

Regenerate your API token from cloud.zilliz.com and update ZILLIZ_CLOUD_TOKEN in claude_desktop_config.json. Tokens may be scoped to specific projects — ensure yours has the correct permissions.

uvx zilliz-mcp-server fails with 'package not found'

Ensure uv is up to date: 'pip install -U uv'. If the PyPI package name differs from the GitHub repo, try cloning the repo and running 'uv run src/zilliz_mcp_server/server.py' instead.

Search returns empty results despite data being inserted

After inserting data, Milvus/Zilliz may require a flush or index build before search becomes effective. Ask Claude to 'flush the documents collection' or wait a few seconds before searching, especially on free-tier clusters.

Frequently Asked Questions about Zilliz MCP

What is Zilliz MCP?

Zilliz MCP is a Model Context Protocol (MCP) server that enables ai agents to interact with milvus vector databases and zilliz cloud through natural language, allowing users to create clusters, manage collections, insert vector data, and perform semantic searches directly from their ai assistants. It connects AI assistants to external tools and data sources through a standardized interface.

How do I install Zilliz MCP?

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

Which AI clients work with Zilliz MCP?

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

Is Zilliz MCP free to use?

Yes, Zilliz MCP 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": { "zilliz-mcp-server": { "command": "npx", "args": ["-y", "zilliz-mcp-server"] } } }

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

Read the full setup guide →

Ready to use Zilliz MCP?

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