PostHog
Open source product analytics platform with feature flags, session replay and A/B testing.
Overview
PostHog started as a product analytics tool and has grown into a broader product-development platform: alongside event-based and web analytics, it includes session replay, feature flags, A/B experiments, error tracking, log ingestion, no-code surveys, a data warehouse that syncs external sources (like Stripe or Hubspot), configurable data pipelines to 25+ destinations, AI observability for LLM-powered apps, and automation workflows - all viewable from one product.
It's installed via a JavaScript snippet, official SDKs (JS, Python, Node, mobile and other backend languages), or its API, and can also be connected as an MCP server to bring PostHog's data into AI coding agents like Claude Code or Cursor.
The core is MIT-licensed (aside from the separate `ee` directory), and the recommended way to run it is PostHog Cloud (US or EU) with a generous free monthly tier; a self-hosted, open-source "hobby deploy" is also available via Docker, though the maintainers note self-hosted deployments don't get official support and are intended for smaller-scale usage before migrating to Cloud.
- Categories
- Analytics
- Languages
- Python, TypeScript
- License
- MIT
Related Projects
Matomo
GPL-3.0Open source web analytics platform giving full control and ownership of visitor data.
- Analytics
Plausible Analytics
AGPL-3.0Lightweight, open source and privacy-friendly alternative to Google Analytics.
- Analytics
Umami
MITSimple, fast, privacy-focused web analytics alternative that you can self-host.
- Analytics
Beam
Apache-2.0Unified programming model for defining batch and stream data processing pipelines that can run portably across multiple execution engines.
- Analytics
Spotted an error? Suggest an edit on GitHub.