ClickHouse MCP Server

ClickHouse MCP servers enable AI models to interact with ClickHouse databases, providing capabilities for high-performance analytical queries, columnar data processing, and real-time data analytics.

{
  "mcpServers": {
    "mcp-clickhouse": {
      "command": "uv",
      "args": [
        "run",
        "--with",
        "mcp-clickhouse",
        "--python",
        "3.13",
        "mcp-clickhouse"
      ],
      "env": {
        "CLICKHOUSE_HOST": "<clickhouse-host>",
        "CLICKHOUSE_PORT": "<clickhouse-port>",
        "CLICKHOUSE_USER": "<clickhouse-user>",
        "CLICKHOUSE_PASSWORD": "<clickhouse-password>",
        "CLICKHOUSE_SECURE": "true",
        "CLICKHOUSE_VERIFY": "true",
        "CLICKHOUSE_CONNECT_TIMEOUT": "30",
        "CLICKHOUSE_SEND_RECEIVE_TIMEOUT": "30"
      }
    }
  }
}

Overview

The MCP ClickHouse Server enables AI models to interact with ClickHouse databases, providing a standardized interface for working with this high-performance columnar database management system. ClickHouse is designed for online analytical processing (OLAP) and is capable of processing billions of rows and terabytes of data in real-time. 1

Key features include:

  • Execute high-performance analytical queries
  • Work with columnar data structures
  • Process real-time analytics workloads
  • Handle large-scale data aggregations
  • Perform complex data analysis

Tools

Sources

Footnotes

  1. https://github.com/ClickHouse/mcp-clickhouse