Skip to content
OpenSourceSoftware.ai

GitHub stars

39,734
—Awaiting first sync

Awaiting first sync

No measured star history yet

What is Umami?

Umami is a lightweight alternative to cookie-based analytics. It brings traffic, campaigns, behaviour, conversions and revenue into one dashboard while avoiding cookies and surveillance, and it can run on your own server or in Umami's cloud.

Umami is an open-source, privacy-first web analytics platform that tracks traffic, campaigns, behaviour, conversions and revenue without cookies. It runs on Node.js with a PostgreSQL database or from the official Docker images, and is also available as a hosted cloud service.

Core capabilities

  • Track traffic, campaigns, conversions and revenue without cookies or surveillance.
  • Self-host with Node.js and PostgreSQL, or run the official Docker images.
  • Connect AI tools through an optional MCP endpoint secured with API keys.

Understand visitors without tracking them

Umami collects the traffic, campaign, behaviour, conversion and revenue data a site owner needs while avoiding cookies and surveillance. Reports stay useful for marketing decisions without building profiles of individual visitors.

  • Privacy
  • No cookies
  • Conversions

Run it on your own infrastructure

Umami needs a server with Node.js 18.18 or newer and a PostgreSQL database, or it can be deployed with the official Docker image and Docker Compose file. Optional settings add two-factor authentication and an MCP endpoint for AI tools that authenticate with API keys.

  • Docker
  • PostgreSQL
  • MCP
  • Two-factor auth

Where it fits

Use cases

  1. 01

    Measure a website without a cookie banner

    Add the Umami script to a site to see traffic and visitor behaviour without setting cookies, so people are not tracked across the web.

  2. 02

    Report on campaigns and conversions

    Follow campaigns, conversion goals and revenue in one dashboard to see which channels bring visitors who buy or sign up.

  3. 03

    Host analytics for several client sites

    Run one self-hosted Umami instance with a PostgreSQL database and add each client website to it, keeping all the data on your own server.