guide

LaunchDarkly Pricing, Decoded - What It Really Costs and Why the Estimate Lies (2026)

LaunchDarkly bills on service connections plus client-side MAU, charged on your highest-volume context kind. Here's how the model actually works, what real contracts cost, and how to estimate your own bill.

Published:

Most “LaunchDarkly pricing” pages either recite the marketing tiers or throw up their hands. Neither helps, because the reason people search this term is that the sticker math misleads you - the number you estimate from your user count and the number on the invoice are two different numbers, and the gap is where the frustration lives.

This is a neutral decode of the actual model. What LaunchDarkly bills on, why the estimate lies, what real contracts cost according to third-party data, and how to work out your own number before you talk to sales. No affiliate angle - just the mechanics nobody writes straight.

What LaunchDarkly bills on

Start with what it does not bill on: seats. Seats are unlimited on every tier, including the free one. You are never taxed for adding a developer. Good - but that’s not where the cost is.

The paid Foundation tier bills on two meters:

  • Service connections - $10 per connection per month. A service connection is roughly an environment-plus-SDK-type pairing your app uses to talk to LaunchDarkly.
  • Client-side monthly active users - $8.33 per 1,000 MAU.

The Developer tier is $0 (core targeting, unlimited seats, ~38 client SDKs), and Enterprise is contact-sales, adding advanced targeting, release automation, workflows, 100-day retention and guarded releases - the full capability breakdown is on our LaunchDarkly review. But the Foundation meters above are the model you need to understand, because the MAU line is the sharp edge.

TierPriceWhat you pay on
Developer$0Core targeting, unlimited seats, no card stated
FoundationUsage-based, billed yearly$10/service connection + $8.33/1k client-side MAU
EnterpriseContact salesCustom, plus advanced targeting and release automation

The part that makes the estimate lie: highest-volume context kind

Here’s the mechanic that surprises people. LaunchDarkly counts monthly active contexts, and bills on the single context kind with the highest volume.

A “context” isn’t just a logged-in user. It can be a user, a device, an anonymous session, an organization, or any custom kind you define. LaunchDarkly tallies each kind separately and charges you on whichever one is biggest.

Work the canonical example. Say you track 300,000 monthly users and 500,000 monthly devices. You’d naturally estimate your bill from the 300,000 users, because that’s what you target flags against. But devices are the higher-volume kind, so you’re billed on 500,000 MAU - even if you never once target a device. At $8.33 per 1,000, that’s the difference between budgeting for 300k and paying for 500k, a two-thirds miss before you’ve added anything.

And devices routinely dwarf users. One user with a phone, a laptop and a tablet is three devices. Anonymous pre-login sessions count too, and a high-traffic marketing site can generate far more anonymous contexts than registered users. This is the single reason a careful headcount-based estimate comes in low. You have to count your highest-volume kind, not your user count.

What real contracts actually cost

LaunchDarkly doesn’t publish enterprise contract prices, so the honest source here is third-party buyer data. According to Vendr, which aggregates real signed contracts, the median LaunchDarkly deal is around $72,000 a year, with a range from $19,500 to $165,700. That median works out to roughly $6,000 a month, and it’s a usage-based figure, not a fixed one - where you land in that range depends almost entirely on your context volume and add-ons.

To be clear about attribution: those numbers are Vendr’s, drawn from buyer contracts, not from LaunchDarkly’s pricing page. Treat them as a reality check on the sticker math, not as a quote.

Add-on stacking

The base license is only part of the bill. Capabilities layer on top as add-ons - experimentation, for instance, is gated to higher tiers and priced separately rather than included. These compound, and at scale they can overtake the base entirely: at 10 million MAU, add-ons alone can exceed $33,700 a month - more than the base license itself.

So when you model LaunchDarkly, you’re modeling at least three things: service connections, MAU on your highest-volume kind, and whichever add-ons you actually need. The last one is easy to under-count because it’s where the advanced features you’re buying LaunchDarkly for often live.

The renewal jump

The recurring theme in buyer reports is the renewal. Once you’re integrated - flags wired through your codebase, SDKs deployed, releases depending on the platform - switching cost is high, and pricing reflects that. Buyers report proposed renewals at roughly double their current cost, particularly when they’re on what LaunchDarkly calls a “legacy plan” and get moved to current pricing.

None of this is unique to LaunchDarkly - usage-based enterprise software renews upward as a rule. But it’s worth planning for specifically here, because the combination of a usage meter that grows with your traffic and a renewal that resets to current pricing means year two often lands well above year one. Get renewal terms in writing before you sign, and model your context growth, not just your current volume.

How to estimate your own bill

Here’s a practical sequence to get a real number instead of a misleading one:

  1. Count your monthly active contexts per kind, separately. Users, devices, anonymous sessions, organizations, any custom kinds. Don’t sum them - LaunchDarkly bills on the biggest single kind, so find your maximum.
  2. Multiply that maximum by $8.33 per 1,000. If your highest kind is 500,000, that’s 500 x $8.33 = about $4,165/mo on MAU alone.
  3. Add service connections at $10 each per month. Count your environment-and-SDK-type pairings.
  4. Add the add-ons you actually need - experimentation and other higher-tier capabilities are priced on top, not included.
  5. Sanity-check against Vendr’s range. If your estimate lands wildly below the $19,500 to $165,700 band for a team your size, you’ve probably under-counted your highest-volume context kind. That’s the usual error.

The output of that exercise is the number that matters. If it’s comfortable and you need the depth, LaunchDarkly is the best product in the category and worth it. If the MAU line alarms you, the next section is for you.

Cheaper alternatives, and why

If the estimate came back higher than you want, the fix depends on what you can give up. Three tools attack the cost from different directions:

  • Flagsmith - if you want open source and no MAU meter. It bills on requests, not MAU, and it’s genuinely self-hostable under BSD-3-Clause. The cloud free tier is 50,000 requests a month; paid Start-Up is $40/mo on annual billing for 1M requests - the cheapest paid entry in the category. Self-host the open-source build and the software is free outright; you pay in DevOps time to run the Django and Postgres stack.
  • ConfigCat - if you want flat, headcount-independent pricing. It charges no per-seat and no per-MAU fees at all, with unlimited MAUs, contexts and flag reads on every tier. You pay by config-download volume instead - Forever Free at $0, Pro at $110/mo, Smart at $325/mo. Just cache your SDK polling, because config downloads are the meter that does move.
  • Statsig - if you want flag checks to be free. It never bills flag or config evaluations; they’re unlimited and free on every tier. It monetizes analytics events instead, so pure flag usage is close to free, with a flat $150/mo Pro tier. The one asterisk is ownership - OpenAI acquired Statsig in September 2025, so factor roadmap uncertainty into a multi-year bet.

What none of these fully match is the top of LaunchDarkly’s feature set - the targeting depth, guarded releases with automatic rollback, and around 38 SDKs. That’s the actual trade you’re pricing: capability versus cost.

The bottom line on pricing: LaunchDarkly isn’t overpriced for what it does. It’s mispriced by you if you estimate from headcount instead of your highest-volume context kind. Count the right number first, add the add-ons honestly, plan for the renewal, and then decide - because that number, not the feature list, is what determines whether LaunchDarkly is the right call.


Pricing verified against launchdarkly.com on 23 July 2026. Feature-flag pricing changes often - we re-verify regularly. Contract figures ($72,000 median, $19,500 to $165,700 range) are from Vendr’s third-party buyer data and attributed as such, not from LaunchDarkly’s published pricing.

Frequently Asked Questions

How much does LaunchDarkly cost per month?

There's no simple sticker price - it's usage-based. The Foundation tier bills $10 per service connection per month plus $8.33 per 1,000 client-side monthly active users, on your highest-volume context kind. Real contracts, per Vendr's buyer data, run a median around $72,000 a year, which is roughly $6,000 a month, ranging from $19,500 to $165,700 annually. The free Developer tier is $0 with unlimited seats.

What is a context kind in LaunchDarkly pricing?

A context is any entity you target a flag against - a user, a device, an anonymous session, an organization. LaunchDarkly counts monthly active contexts and bills you on the single kind with the highest volume. So if you have 300,000 monthly users but 500,000 monthly devices, you're billed on 500,000, the larger number, even if you only ever target users. This is the detail that makes headcount-based estimates wrong.

Why is my LaunchDarkly renewal so much higher?

Renewals are the most-cited complaint. Buyers report proposed renewals at roughly double their current cost, particularly when moving off a "legacy plan." Combined with add-ons that stack on top of the base license - which at 10M MAU alone can exceed $33,700 a month - the second-year bill often lands well above the first. Model your context growth before you sign, and get renewal terms in writing.

Is LaunchDarkly worth the price?

For large engineering orgs that need the deepest targeting, guarded releases with automatic rollback, and around 38 SDKs, yes - nothing else matches the depth, and its G2 score of 4.5 across 681 reviews reflects that. For cost-sensitive teams, or ones with high device or anonymous-context counts, the usage-based model often makes a cheaper option like ConfigCat, Flagsmith or Statsig the better call. The deciding factor is whether you need the top of the feature set.

Explore More

Free Newsletter

Get the Feature Flags Newsletter

Platform benchmarks, real pricing data and progressive delivery practice. No spam.

Free. Unsubscribe any time. See our privacy policy.

Related Articles