---
title: "Zilliz MCP Server: review, setup & limitations · Undominated.ai"
canonical: https://undominated.ai/mcp-servers/zilliz-cloud-mcp/
description: "Manages Zilliz Cloud projects and clusters and works with data in Milvus collections (create, insert, delete, vector and hybrid search) from an MCP client."
---

# Zilliz MCP Server: review, setup & limitations · Undominated.ai

> Manages Zilliz Cloud projects and clusters and works with data in Milvus collections (create, insert, delete, vector and hybrid search) from an MCP client.

[← Explore all mcp servers](/mcp-servers/)

VECTOR DATA MANAGEMENT / Zilliz

# Zilliz MCP Server

Manages Zilliz Cloud projects and clusters and works with data in Milvus collections (create, insert, delete, vector and hybrid search) from an MCP client.

 [See setup guidance ↓](#setup)[Original source ↗](https://github.com/zilliztech/zilliz-mcp-server)

SOURCE REVIEW

 Reviewed 2026-10-13
 Evidence 5 linked sources
 Publisher Zilliz
 Licence [Apache-2.0 ↗](https://github.com/zilliztech/zilliz-mcp-server/blob/c2a84e1779d87aa2ba7098b6609f8a3cb529c4e6/LICENSE)
 Revision c2a84e1779d8
 [Read what was—and wasn’t—checked ↓](#review)

“MCP server for Zilliz Cloud and Milvus”

 [Zilliz · upstream description ↗](https://github.com/zilliztech/zilliz-mcp-server/blob/c2a84e1779d87aa2ba7098b6609f8a3cb529c4e6/pyproject.toml) Our analysis follows below.

01 / THE REASONING

## Why this made the selection.

 - Published in Zilliz's own repository; the README at this commit separates control-plane tools (list_projects, list_clusters, create_free_cluster, describe_cluster, suspend_cluster, resume_cluster, query_cluster_metrics) from data-plane tools on Milvus collections.
- pyproject.toml names zilliz-mcp-server 1.0.0 with license Apache-2.0, matching the LICENSE file, and declares the zilliz-mcp-server console script the README's uvx block runs.

### A good fit for

 - Provisioning a Zilliz Cloud cluster and trying vector searches against it from an agent
- Checking a cluster's performance metrics and a collection's schema with query_cluster_metrics and describe_collection

### Weigh up before choosing

 - Data-plane tools create collections and insert or delete entities; control-plane tools create, suspend and resume clusters.
- settings.py reads the credential from ZILLIZ_CLOUD_TOKEN and refuses to start without it, but the README's Streamable HTTP instructions put ZILLIZ_API_KEY in .env, so following them as written would leave the token unset.
- Zilliz's separate mcp-server-milvus is already recorded as milvus. Nothing was run.

02 / THE REVIEW RECORD

## What we actually inspected.

Source review has boundaries. A clear record is more useful than a “safe” badge.

### Material inspected

 - README.md at c2a84e1779d87aa2ba7098b6609f8a3cb529c4e6
- LICENSE, pyproject.toml, src/zilliz_mcp_server/settings.py and server.py at the same commit

### Our findings

 - server.py takes --transport (default stdio); settings.py defaults ZILLIZ_CLOUD_URI to https://api.cloud.zilliz.com and the free-cluster region to gcp-us-west1, and loads .env only without overriding existing variables.
- The README's stdio block carries a note about replacing /path/to/your/zilliz-mcp-server that the block itself does not use.

### Not established by this review

 - uvx launch
- Any Zilliz Cloud or Milvus call
- create_free_cluster

The review applies to the material and revision named here. A newer upstream release can change its behavior.

03 / PUT IT TO WORK

## Connect a server deliberately.

[Upstream setup instructions ↗](https://github.com/zilliztech/zilliz-mcp-server/blob/c2a84e1779d87aa2ba7098b6609f8a3cb529c4e6/README.md)
 - The README gives a client block rather than a one-line command: command uvx with args zilliz-mcp-server and ZILLIZ_CLOUD_TOKEN in env. It was not run.
- For HTTP mode, set ZILLIZ_CLOUD_TOKEN (the variable settings.py reads) before running the README's uv run command. Do not commit the token.

### Before you start

 - uv and Python 3.10 or later
- A Zilliz Cloud account and API key

### Documented configuration example

Use the client and authentication context described in the setup guidance. Configuration formats vary between hosts.

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

### Compatibility

Claude · Cursor

### Implementation

Python

### Transports

stdio · Streamable HTTP

### Local process with a Zilliz Cloud API key

 - Lists projects and clusters and creates, suspends or resumes clusters in the Zilliz Cloud account
- Creates collections, inserts, deletes and searches entities in Milvus databases the key can reach

### Cost model

Apache-2.0 server source. Zilliz Cloud clusters beyond the free tier are billed by Zilliz; no rate was reviewed.

04 / FOLLOW THE EVIDENCE

## The source trail.

Our notes are separate from the original resource. Check upstream before adopting a new version.

 - [Zilliz MCP README ↗](https://github.com/zilliztech/zilliz-mcp-server/blob/c2a84e1779d87aa2ba7098b6609f8a3cb529c4e6/README.md) Checked 2026-10-13 https://github.com/zilliztech/zilliz-mcp-server/blob/c2a84e1779d87aa2ba7098b6609f8a3cb529c4e6/README.md Supports: Tools, stdio and Streamable HTTP setup
- [Zilliz MCP settings.py ↗](https://github.com/zilliztech/zilliz-mcp-server/blob/c2a84e1779d87aa2ba7098b6609f8a3cb529c4e6/src/zilliz_mcp_server/settings.py) Checked 2026-10-13 https://github.com/zilliztech/zilliz-mcp-server/blob/c2a84e1779d87aa2ba7098b6609f8a3cb529c4e6/src/zilliz_mcp_server/settings.py Supports: ZILLIZ_CLOUD_TOKEN requirement, Defaults
- [Zilliz MCP server.py ↗](https://github.com/zilliztech/zilliz-mcp-server/blob/c2a84e1779d87aa2ba7098b6609f8a3cb529c4e6/src/zilliz_mcp_server/server.py) Checked 2026-10-13 https://github.com/zilliztech/zilliz-mcp-server/blob/c2a84e1779d87aa2ba7098b6609f8a3cb529c4e6/src/zilliz_mcp_server/server.py Supports: Transport flag
- [Zilliz MCP pyproject.toml ↗](https://github.com/zilliztech/zilliz-mcp-server/blob/c2a84e1779d87aa2ba7098b6609f8a3cb529c4e6/pyproject.toml) Checked 2026-10-13 https://github.com/zilliztech/zilliz-mcp-server/blob/c2a84e1779d87aa2ba7098b6609f8a3cb529c4e6/pyproject.toml Supports: Description, Package 1.0.0, License Apache-2.0
- [Zilliz MCP LICENSE ↗](https://github.com/zilliztech/zilliz-mcp-server/blob/c2a84e1779d87aa2ba7098b6609f8a3cb529c4e6/LICENSE) Checked 2026-10-13 https://github.com/zilliztech/zilliz-mcp-server/blob/c2a84e1779d87aa2ba7098b6609f8a3cb529c4e6/LICENSE Supports: Apache License 2.0

KEEP COMPARING

## Other approaches to consider.

Related by category or shared topics. These are alternatives to inspect, not a measured quality order.

 [### Chroma MCP ↗ Lets an assistant create and query Chroma collections through local, HTTP or Chroma Cloud database connections.](/mcp-servers/chroma/)[### Pinecone MCP ↗ Search Pinecone documentation and manage or query indexes that use integrated embeddings.](/mcp-servers/pinecone/)[### Weaviate built-in MCP ↗ Inspect collections, search vector and keyword data, and optionally upsert objects through Weaviate’s built-in MCP endpoint.](/mcp-servers/weaviate/)

 [AI Tools ↗](/tools/)[Skills ↗](/skills/)[Agents ↗](/agents/)[MCP Servers ↗](/mcp-servers/)[Workflows ↗](/workflows/)

## Continue your investigation

 - [Inspect resource evidence](/tools/)
- [Choose a host](/agents/)
- [Inspect reusable workflows](/skills/)
- [Use Undominated data](/api/)
