PostHog Feature Flags Pricing in 2026 - Why the Free Million Is the Cheap Part

PostHog’s flag pricing looks almost too good on the pricing page. A million feature-flag requests a month, free, no credit card. For a lot of teams that is the whole bill - zero. So why does anyone search “PostHog feature flags pricing” worried about cost?

Because the flags line is not where PostHog makes money, and it is not where your bill grows. This is a neutral decode of how the flag meter actually works, the worked math on what you would pay past the free tier, and the multi-product trap that catches people who think of PostHog as a flag tool. It is an analytics suite first, and that framing is the whole pricing story.

What PostHog bills flags on

Feature flags are metered per request, and the first million every month are free. After that it is a descending tier ladder - the more requests you send, the less each one costs.

Monthly flag requests	Price per request
First 1,000,000	Free
1M to 2M	$0.0001
2M to 10M	$0.000045
10M to 50M	$0.000025
50M+	$0.000010

There is no base fee for the flags line - it is purely usage-based after the free tier, and there is no per-seat or per-MAU charge anywhere. Experiments and A/B tests draw from the same meter and the same 1M free requests, so running tests does not open a separate bill. For flags alone, this is one of the cheapest models in the category. The PostHog review has the full breakdown.

The trap - each product bills separately

Here is the part the flag pricing hides. PostHog is a product-analytics suite - events, funnels, session replay, surveys, a data warehouse - with flags as one product line among many. And every product meters and bills on its own.

Your flags line can sit inside the free million all year while your analytics events, session replays and surveys each rack up their own usage-based charges. Teams that adopt PostHog “for the cheap flags” and then switch on the rest of the suite are surprised when the invoice arrives, because they were watching the wrong meter. The flags stayed cheap. The platform did not.

Model every product you plan to turn on, not just flags. That is the single mistake that makes PostHog feel expensive.

Estimate your flag bill

If you want a real number for the flags line specifically, the math is easy because the tiers are published. Say your app sends 10 million flag requests a month:

1. First 1,000,000 requests - free.
2. Next 1,000,000 (the 1M to 2M tier) at $0.0001 = $100.
3. Next 8,000,000 (the 2M to 10M tier) at $0.000045 = $360.
4. Total flags bill - about $460/mo.

No seats, no MAU, no base fee added. Now do the honest version of the exercise - repeat it for every other product you will use. Analytics events and session replays each have their own meter and their own free allowance, and those are where a full-suite PostHog bill actually lives. The flags number above is often the smallest line on the invoice.

One more thing the low flag price hides - the open-source self-host path is officially unsupported. PostHog’s main product is MIT-licensed, but self-hosted deployments come with no guarantees, no tagged releases and you assume the risk. So “run it yourself for free” is not a real way to dodge the meter for most teams. Cloud is the product.

Cheaper alternatives, and why

If the multi-product meters worry you, or you want flags without adopting an analytics suite, two tools attack the cost from different angles.

• Statsig - if you want flag checks to be free and a flat number to budget. Statsig never bills flag or config evaluations - they are unlimited and free on every tier, including the free one. It monetizes analytics events instead. The free Developer tier gives 2M events and 50,000 session replays with no card, and Pro is a flat $150/mo rather than a per-request meter. The asterisk is ownership - OpenAI acquired Statsig on 2 September 2025 for a reported ~$1.1 billion, so a multi-year bet on it rides on a roadmap that now sits inside OpenAI.

• Flagsmith - if you want flags only, self-host, and no lock-in. Flagsmith is flags-first, open source under BSD-3-Clause, and genuinely self-hostable via Docker and Kubernetes. The cloud free tier is 50,000 requests a month with unlimited flags, and paid Start-Up is $40/mo on annual billing for 1M requests - the cheapest paid entry in the category. Self-host the OSS build and the software is free outright; you pay in DevOps time to run the Django and Postgres stack.

What neither of those gives you is PostHog’s real advantage - experiments that read directly from your product analytics, so the metrics you test against are the same events you already track. That coupling is the reason to pay PostHog anything at all.

The bottom line on PostHog pricing: the flags are genuinely cheap, and the free million is real. But you are not pricing a flag tool - you are pricing an analytics suite where flags happen to be the cheapest product. Count every meter you will turn on, watch analytics and replay rather than flags, and PostHog stays the good deal it looks like. Ignore the other products and the bill will surprise you exactly the way LaunchDarkly’s renewals surprise people, just from the opposite direction.

---

Pricing verified against posthog.com on 26 July 2026 via our PostHog review. Feature-flag pricing changes often - we re-verify regularly. The Statsig acquisition figure (~$1.1 billion, 2 September 2025) is from third-party reporting and attributed as such.