MCP
Connect Claude, ChatGPT, Cursor, and other AI tools to your Umami Cloud analytics using the Model Context Protocol.
Umami Cloud provides a remote Model Context Protocol (MCP) server. It lets AI tools like Claude, ChatGPT, and Cursor answer questions about your website traffic in natural language.
Self-hosting Umami? See MCP for self-hosted setup.
All tools are read-only. The MCP server calls the same Cloud API as the dashboard, so it respects your existing website and team permissions. It never accesses your data directly.
Connect#
The MCP endpoint is:
Authenticate with an API key. Pass it in the Authorization header using the Bearer scheme:
Clients that support custom headers can use x-umami-api-key instead. If both headers are provided, they must contain the same key.
Use an MCP client that supports API-key or bearer-header configuration. The same subscription requirements and rate limits as the Cloud API apply.
Client setup#
Most MCP clients accept a remote server URL and a set of headers. For example, in Claude Desktop or Cursor, add the following to your MCP configuration:
If your client only supports local stdio servers, you can run the @umami/mcp package with npx and point it at Umami Cloud using your API key:
Tools and example prompts#
Cloud and self-hosted instances expose the same tools. See MCP for the full tool list, example prompts mapped to each tool, and limitations.
Plan requirements#
MCP uses Cloud API keys, which are available on the Pro, Business, and Enterprise plans. See pricing.
Revoking access#
Access is tied to the API key. To disconnect an AI tool, delete the key under Settings → API keys.