GrowthBook Pricing, Decoded - Cloud Seats vs Free Self-Host in 2026
GrowthBook charges per seat on Cloud and nothing to self-host, but the real cost is the warehouse and the stats learning curve. Here's how to estimate your bill and when Statsig or PostHog is cheaper.
Published:
If you’re pricing GrowthBook, you’ve probably already worked out that it’s the cheap serious-experimentation option. That’s true. But “cheap” hides a real question - do you pay in per-seat dollars, or in warehouse bills and DevOps time? GrowthBook offers you both paths, and picking the wrong one is where teams overspend.
Here’s the model decoded straight: what the meters are, how to estimate your number, and when Statsig or PostHog works out cheaper.
What GrowthBook bills on
GrowthBook is unusual because it has two completely different pricing paths, and they meter nothing alike.
On Cloud, you pay per seat. No MAU, no flag meter, no event meter - just people. Self-hosted, you pay nothing to GrowthBook at all - the core is MIT-licensed and runs unlimited users for free. What you pay instead is your warehouse’s compute bill and the engineering time to operate the platform. That’s the trade at the heart of every GrowthBook cost decision.
| Tier | Price | What you get |
|---|---|---|
| Starter (Cloud) | $0 | 3 users, 1 project, unlimited flags/experiments/traffic |
| Pro (Cloud) | $40/seat/mo | 50 users, 3 projects, visual editor, bandits, safe rollouts |
| Open Source (self-hosted) | $0 | Unlimited users, 1 project, bring your own warehouse |
| Enterprise | Contact sales | SSO, SCIM, ramp schedules, audit logs, 99.99% SLA |
Both free options are real - Cloud Starter for a small team, and the OSS self-host with unlimited users. Full breakdown on the GrowthBook review.
How to estimate your Cloud bill
Cloud pricing is the simplest math in this whole category, because there’s one variable:
- Count the people who need a seat. Engineers, PMs, analysts, anyone who logs into GrowthBook.
- Under 3? You’re free. Cloud Starter covers 3 users and 1 project at $0.
- Multiply the rest by $40/mo. A team of 10 on Pro is $400/mo. A team of 25 is $1,000/mo. Pro caps at 50 users and 3 projects.
- Need SSO, SCIM, ramp schedules or audit logs? That’s Enterprise, priced on a call.
Because the meter is seats, GrowthBook gets more expensive with people, not traffic. That’s the opposite of event-based rivals, and it’s the key thing to internalize: a low-traffic product with a big cross-functional team can cost more on GrowthBook Cloud than a high-traffic product with three engineers.
The self-host is free software, not a free setup
Here’s the option that reads as “$0” and isn’t quite. The MIT core self-hosts with unlimited users at no license cost. For a large team, that headcount-free math is enormously attractive next to $40 per seat.
But GrowthBook’s whole value is being warehouse-native - it queries the BigQuery, Snowflake or Databricks warehouse you already run rather than ingesting a copy of your events. So two costs replace the license:
- The warehouse itself. No instrumented warehouse, far less value - and that warehouse has its own compute bill every time GrowthBook runs a query. There’s a Managed Warehouse option for teams that don’t want to connect their own, but it adds cost.
- The stats learning curve. CUPED, Bayesian inference, sequential testing and SRM checks are powerful, but they aren’t self-explanatory. Reviews cite an engineer-first UI and thorough-but-hard-to-navigate docs. A team without a data-literate person gets a fraction of the value the feature list promises.
So the free self-host is genuinely free on license and genuinely not free on setup. GrowthBook rewards teams that already have data maturity and punishes teams that don’t. If that’s you, it’s the best-value serious experimentation platform available. If it isn’t, budget for the ramp.
When Statsig or PostHog is cheaper
GrowthBook’s rivals bill on entirely different meters, so the cheaper option depends on your shape:
- Statsig - if you want flat, predictable pricing and free flags. Statsig never bills flag or config checks - they’re unlimited and free on every tier. It monetizes analytics events instead. The free Developer tier gives 2M events; Pro is a flat $150/mo with 5M events, then $0.05 per 1,000. Because it’s flat rather than per-seat, a big team pays the same $150 that a small one does - so for a large cross-functional team, Statsig can undercut GrowthBook Cloud outright. Its experimentation engine, built by ex-Facebook engineers, is the strongest here. The Statsig pricing decode has the full model, and Statsig vs GrowthBook weighs them directly.
- PostHog - if you want experiments bundled with analytics and don’t have a warehouse. PostHog’s experiments read straight from its own product analytics, so you don’t need a separate instrumented warehouse the way GrowthBook does. Flags and experiments share one meter - the first 1M requests a month are free, then fractions of a cent. The catch is that each PostHog product bills separately, so analytics, replay and surveys each add to the total even while the flags line stays cheap. See the PostHog pricing decode and GrowthBook vs PostHog.
The bottom line on GrowthBook pricing: the sticker is honest and the free tiers are real. The number that actually decides your cost isn’t on the pricing page - it’s whether you already have an instrumented warehouse and someone to drive the stats. Have both, and self-hosting is close to unbeatable value. Have neither, and a flat-priced or analytics-bundled rival will serve you cheaper and faster. For the wider field, our GrowthBook alternatives and best experimentation platforms guides map the rest.
Pricing verified against growthbook.io on 26 July 2026. Feature-flag pricing changes often - we re-verify regularly. Funding figures are third-party-sourced and labelled as such.
Frequently Asked Questions
How much does GrowthBook cost?
Cloud Starter is $0 for up to 3 users and 1 project with unlimited flags, experiments and traffic. Cloud Pro is $40 per seat per month for up to 50 users and 3 projects, adding the visual editor, multi-arm bandits and safe rollouts. The self-hosted open-source edition is $0 with unlimited users and 1 project. Enterprise is contact-sales and adds SSO, SCIM, ramp schedules and audit logs. There's no published annual discount.
Is GrowthBook really free to self-host?
Yes, the software is. The open-source core is MIT-licensed and self-hosts with unlimited users at no license cost. But the value comes from querying your own data warehouse, so the real costs are the warehouse you connect and the DevOps time to run the platform. Several directories in the repo sit under a separate commercial license, and features like SSO and SCIM are gated to Enterprise. Free software, not a free setup.
What does warehouse-native mean for the price?
GrowthBook queries the BigQuery, Snowflake or Databricks warehouse you already run instead of ingesting a copy of your events. That means it doesn't bill you on event volume the way Statsig or PostHog do - your analytics data and its compute cost stay in your warehouse. The trade is that you need an instrumented warehouse for GrowthBook to deliver value, and that warehouse has its own bill.
Is GrowthBook cheaper than Statsig?
For a small team, both have generous free tiers, so it's a wash at the start. GrowthBook Pro is $40 per seat, so cost scales with headcount; Statsig Pro is a flat $150/mo and bills on analytics events, not seats or flags. If you have many users touching the tool, GrowthBook's per-seat model adds up. If you send heavy event volume, Statsig's event meter is the one to watch. Different meters, so estimate against your own shape.
Explore More
Tool Reviews
Related Articles
- PostHog Experiments Pricing, Decoded - Why "Free A/B Testing" Has a Second Meter (2026)
- Statsig Pricing, Decoded - Why Flags Are Free and the Bill Lives in Events (2026)
- The Feature Flag Consolidation Map - Who Got Bought in 2024 to 2026
- How to Choose Metrics for A/B Testing - Primary, Secondary and the OEC (2026)
- What Is a Holdout Group? Measuring the Cumulative Impact of Experiments (2026)
Free Newsletter
Get the Feature Flags Newsletter
Platform benchmarks, real pricing data and progressive delivery practice. No spam.
Related Articles
Bayesian vs Frequentist A/B Testing - Which Stats Engine to Trust (2026)
Your A/B tool answers a Bayesian question or a frequentist one, and they are not the same question. Here is what each actually computes, how to read the output correctly, and which platforms use which.
July 28, 2026
guideConfidence Intervals in A/B Testing - How to Read Them Right (2026)
A confidence interval tells you the plausible range of your true lift, which is more useful than a pass-fail p-value. Here is how to read one, the overlap trap, and relative vs absolute lift.
July 28, 2026
guideCUPED Variance Reduction in A/B Testing, Explained (2026)
CUPED uses pre-experiment data to cut the noise in your metrics, so tests reach significance on less traffic. Here is how CUPED works, the intuition and the math, when it helps most, and which platforms support it.
July 28, 2026
GrowthBook Review
Statsig Review
PostHog Review