GrowthBook
Open-source feature flags plus warehouse-native A/B testing.
About
GrowthBook bundles feature flags, statistical A/B testing, and product analytics into one warehouse-native platform, with the entire stack open source under MIT and self-hostable for free. Its experimentation engine runs against your own data warehouse, and the stats are fully inspectable rather than a black box. The catch is that the warehouse-native model means you need reliable event or warehouse data to get value, and production self-hosting is your team's operational responsibility.
What it can do
- Feature flags with safe rollouts and guardrails
- Warehouse-native A/B testing with open stats engine
- Product analytics with funnels and dashboards
- Multi-armed bandits and visual experiment editor
- REST API and MCP server endpoints
Setup
- Self-host the open-source edition or sign up for Cloud Starter (free)
- Connect your data warehouse as the metrics source
- Install an SDK to add feature flags
- Launch experiments and read results in the warehouse
Prerequisites
A data warehouse (BigQuery, Snowflake, etc.) or use managed events
Use cases
Works with
Trust score — 8/10
Scored 8/10 (verified tier, metrics-v1). Evidence: MIT licensed, ~8,175 GitHub stars as of Aug 2026; Pricing page and multiple 2026 comparisons corroborate plans; Warehouse-native model requires own data infra. License: MIT (open source); cloud plans are proprietary SaaS..
★ 8,175 GitHub stars
How the trust score is built
Full methodology →- Adoption · stars/downloads, log-scaled vs peers
- 9.0/10
- Maintenance · recency, releases, issue load
- 9.0/10
- Openness · license + docs
- 10.0/10
- Manual review · sanity read + red-flag check
- 7.0/10
- · MIT licensed, ~8,175 GitHub stars as of Aug 2026
- · Pricing page and multiple 2026 comparisons corroborate plans
- · Warehouse-native model requires own data infra
Sources
Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.