PostHog
Product data platform combining analytics, session replay, feature flags, error tracking and a warehouse, with AI on top.
PostHog is a product data platform that combines analytics, session replay, feature flags, experiments, error tracking and a data warehouse. PostHog is used by installing its SDK or setup wizard in an application, after which teams explore the data in the web app, with SQL, or through PostHog AI and an MCP server for agents.
Best for Engineering and product teams that want analytics, replays, flags and error data in one system their AI agents can also query.
- Platforms
- Web, API, MCP
- Pricing model
- Freemium
- Free plan
- Yes
- API
- Yes
- Made by
- PostHog
- Added
Last verified: . Information taken from posthog.com.
What is PostHog?
PostHog is a product data platform that collects analytics events, session recordings, errors, logs and business data in one place so that teams and AI agents can query them together. It bundles a long list of tools that are usually bought separately: product and web analytics, session replay, feature flags, experiments, surveys, error tracking, a managed data warehouse with a SQL editor, a customer data pipeline, workflows and AI observability. Data from more than 120 outside sources, such as Stripe, Postgres and HubSpot, can be synced into the warehouse. On top of that sits PostHog AI, which answers questions across the combined data, an MCP server for coding agents, and an Inbox feature in beta that groups findings into reports and drafts pull requests. Pricing is usage-based per product with a monthly free allowance, and an open-source edition is published under the MIT license.
Key Features
Product and Web Analytics
Tracks events, funnels and traffic, with one million events included free each month.
Session Replay
Records user sessions for playback and detects signals such as dead clicks and long stalls.
Feature Flags and Experiments
Controls rollouts and runs A/B tests, billed together on flag requests.
Error Tracking
Groups exceptions and stack traces into issues that can be reviewed alongside recordings.
Data Warehouse
Stores synced rows from external sources and exposes them through a SQL editor, BI views and an API.
PostHog AI and MCP
Lets people ask questions in plain language and gives coding agents direct access to product data from the editor.
Use Cases
- Diagnosing a Conversion Drop
- Combine funnel data, payments and recent code changes to find where users abandon checkout.
- Gradual Feature Releases
- Ship a change to a share of users and compare results before a full launch.
- Bug Triage
- Watch the session that produced an exception to understand what the user did.
- Self-Serve Reporting
- Query product and revenue data with SQL without maintaining a separate warehouse.
- Monitoring LLM Features
- Use AI observability to track events generated by an application's model calls.
Things to know
- Free tier resets monthly
- Allowances such as 1M analytics events, 5K recordings and 100K exceptions renew each month, and unused amounts do not carry over.
- Free plan limits
- Without a card the account has one project, one-year data retention and community support.
- Billing limits
- On pay-as-you-go, a spending cap can be set per product; data beyond the cap is dropped rather than billed.
- Hosting regions
- PostHog Cloud can be hosted in the US (Virginia) or the EU (Frankfurt).
FAQs
What is PostHog?
PostHog is a platform that gathers product analytics, session replays, errors and business data in one place. Teams and AI agents query that combined data to understand and improve a product.
What is PostHog used for?
PostHog is used to measure how people use software, replay sessions, roll out features behind flags, run experiments and track errors.
Is PostHog free?
Yes. PostHog has a free plan with a monthly allowance on every product, no credit card and no time limit. Usage stops at the free limits unless a card is added.
How much does PostHog cost?
PostHog charges per product for usage above the monthly free allowance, with prices quoted in US dollars. Per-unit rates and a calculator are on the official pricing page.
Does PostHog have an API?
Yes. PostHog lists an API and webhooks as part of its platform, along with an MCP server for AI agents.
Is PostHog open source?
PostHog publishes an open-source version under the MIT license in addition to its hosted cloud product.
How long does PostHog keep data?
PostHog retains events for one year on the free plan and seven years on a paid plan. Session recordings are kept for one month on the free plan and 90 days on pay-as-you-go.
Who it's for
- Product engineers
- Product managers
- Startups
- Data teams
- Growth teams
Used in these workflows
- Designing an app interface with AI: research, screens, prototype, codeStage 7: Usability testing and QA
- Building an AI-powered app: the tool stack, layer by layerStage 8: Evaluation and monitoring
- Launching a product with AI: what to use from positioning to feedbackStage 8: Feedback and analytics
- Which AI tools do you need to build a website?Stage 9: Testing, monitoring and analytics
Each guide lists the stages of one job and the tools that help at each stage. All guides
Comments
No comments yet. If you have used PostHog, tell other readers how it went.
Alternatives to PostHog
| Tool | Free plan | Starting price | API |
|---|---|---|---|
 OpenReplayOpen-source session replay and product analytics suite that can be self-hosted or run as a managed instance. |
Yes | — | Yes |
 AirbyteOpen-source data movement platform that replicates data from 700+ sources and exposes it to analytics and AI agents. |
— | — | Yes |
 LangChainOpen source agent frameworks plus the LangSmith platform for tracing, evaluating and deploying AI agents. |
Yes | — | Yes |
 Vanna AIOpen-source SQL agent framework that lets users query company databases in natural language. |
— | — | Yes |
 naoOpen-source framework for building, testing and deploying analytics agents on your own data stack. |
Yes | — | — |
Listed in the same categories as PostHog. A starting price appears only where we checked it on the maker’s pricing page; a dash means not checked or not published. All Data Analysis & Analytics tools