PostHog is your product's context layer
PostHog ingests and stores your analytics, errors, replays, and business data so you and your agents can query and act on it. It gives you every tool you need to build a successful product, and captures the context agents need to diagnose problems, uncover opportunities, and ship fixes:
Analytics
- Product analytics: Autocapture or manually instrument event-based analytics to understand user behavior. Analyze your data with visualizations or SQL.
- Web analytics: Monitor web traffic and user sessions with a GA-like dashboard. Track conversion, web vitals, and revenue.
- Group analytics: Analyze multi-seat accounts and other groups of users.
Session replay
- Session replay: Watch real user sessions from your website or mobile app to diagnose issues and understand user behavior.
- Replay Vision: Let AI watch your recordings for you and turn them into queryable observations.
- Heatmaps: See where users click, scroll, and move on your site.
Observability
- Error tracking: Track errors, get alerts, and resolve issues to improve your product.
- Logs: Ingest, search, and analyze log data alongside the rest of your product data.
- Tracing: Follow distributed traces straight to the line of code that broke.
Ship and iterate
- Feature flags: Safely roll out features to select users or cohorts.
- Experiments: Test changes and measure their statistical impact on goal metrics. Set up experiments with no code, too.
- Early access features (beta): Let users opt into betas and register interest in upcoming features.
Talk to users
- Surveys: Ask anything with our collection of no-code survey templates, or build custom surveys with our survey builder.
- Support: Run customer support with a chat widget and a unified inbox, next to the product data behind each ticket.
- Workflows: Automate actions and messages to your users, triggered by your product data.
Data
- Warehouse: Collect, store, transform, and query all the context behind your product. Includes a managed warehouse, 120+ sources, and a SQL editor with BI and data visualization.
- Data pipelines: Run custom filters and transformations on your incoming data. Send it to 25+ tools or any webhook in real time, or batch export large amounts to your warehouse.
- Endpoints: Build custom API endpoints powered by your PostHog data.
AI and agents
- AI observability: Capture traces, generations, latency, and cost for your LLM-powered app. Run LLM-as-a-judge evaluations to catch regressions.
- MCP analytics (beta): See how agents actually use your MCP server.
- PostHog AI: Ask your product data anything. PostHog AI builds insights, writes SQL, and finds replays for you.
- Self-driving: Turn signals in your product data (errors, rage clicks, failed queries, and more) into researched reports and pull requests, delivered to your Inbox for you to review and merge.
You can steer it all from Slack, web, desktop (PostHog Desktop), or your own editor via the MCP.
The remote MCP server is https://mcp.posthog.com/mcp. It routes to the correct data region (US or EU) when your human logs in. Install it into Claude Code, Cursor, Codex, VS Code, Zed, or PostHog Desktop with:
npx @posthog/wizard mcp add
The docs are indexed for you at posthog.com/llms.txt, the changelog is plain Markdown at posthog.com/changelog.md, and the setup guides for every MCP client are in the docs.
All of this is free to use with a monthly free tier for each product. No card is required, and 97% of users pay $0. Join the 500,000+ teams already shipping with PostHog by signing up for PostHog Cloud US or PostHog Cloud EU.
Table of Contents
Getting started with PostHog
PostHog Cloud (Recommended)
The fastest and most reliable way to get started with PostHog is signing up for free to PostHog Cloud or PostHog Cloud EU. Your first 1 million events, 5k recordings, 1M flag requests, 100k exceptions, 10GB of logs, and 1500 survey responses are free every month, after which you pay based on usage.
Self-hosting the open-source hobby deploy (Advanced)
If you want to self-host PostHog, you can deploy a hobby instance in one line on Linux with Docker (recommended 4GB memory):
This is the beginning of the README. Read the rest at the original. Read the full README ↗
