Powstania Warszawskiego 15
31-539 Kraków
Apple takes 30 percent as standard, 15 percent for small developers, 26 percent in the European Union from October 1, 2026, and nothing at all on real world services. Google lands most subscriptions at 15 percent, or 10 percent if you use your own checkout. Which of those numbers applies to you is decided by what you sell and where your users are, and it is decided before a line of code is written. Here is the current picture, without the developer jargon.
Updated September 16, 20269 min readRates checked on September 16, 2026. This area is moving quickly, so the sections below give the conditions behind each number and the situation that is still unsettled in the United States.
Everything else in this article explains how those numbers are assigned to your specific product, because the mistake that costs the most money is not paying 30 percent. It is paying 30 percent on something that never had to go through the store at all.
The store's own payment system, on Apple and Google alike. Required for anything digital your app sells inside the app: subscriptions, premium features, content.
A payment provider for everything the store's system must not touch: real world services, bookings, physical goods.
Both app stores apply the same test to every payment in every app: where is the thing being consumed? If it is consumed on the phone, the payment goes through the store's billing, or one of the alternatives the store now permits in your market. If it is consumed in the physical world, the payment stays out of the store's system entirely. Apple's version of this rule is the stricter one, so the article works through Apple first and then shows where Google differs.
That single test sorts almost every business model. A subscription to workout videos is consumed on the phone, so it is in app purchase. A haircut, a training session, a delivered meal, a cleaning visit, all of these are consumed in the real world, so they are paid through a provider like Stripe and Apple takes nothing. Not because Apple is being generous, but because its own guideline 3.1.5(a) requires physical goods and services to be paid outside in app purchase.
So before anyone debates tools, write down what your app charges money for. Each item on that list already has its payment path assigned by the guidelines. The tools follow. If you are still working out what the product itself will contain, that same list is the one we start from in an app discovery workshop.
Guideline 3.1.1 is the one that gets apps rejected. Digital content, features, or subscriptions unlocked inside the app must use Apple's in app purchase system, apart from the link out options Apple now allows on the US and EU storefronts. Routing a premium tier through your own checkout without one of those permissions is among the most common rejection reasons in App Review.
The commission is less brutal than the headline 30 percent, if you use the mechanics:
Enrolling in the Small Business Program takes minutes in App Store Connect, requires no code, and halves your commission for as long as you are under the threshold. Do it before launch rather than after, because it applies from the point of approval and not to what you already sold.
If you cross 1 million dollars of proceeds during a year, the standard rate applies to later sales, and you can qualify again in a future year if proceeds fall back below the threshold.
Everything Apple's system is not allowed to handle runs through a standard payment provider, and in mobile projects that provider is usually Stripe. Bookings, marketplace transactions, payouts to service providers, refunds for cancelled appointments. When we built Evesport, a personal trainer booking platform, all of the payment engineering lived here: charging clients for sessions, paying trainers out, handling cancellations. Apple's commission never entered the picture.
Which payments go through the store and which go through Stripe is something we settle during scoping, not during the build. In fixed price app development it has to be, because the checkout, refunds, payouts and review risk are all priced from that one decision.
There is one exception that surprises even experienced teams. Guideline 3.1.3(d) allows realtime person to person services between two individuals, and Apple's own examples include tutoring, medical consultations, and fitness training, to be paid outside in app purchase. A live video session between one coach and one client can go through Stripe, commission free. The same session streamed to a group cannot. One sentence in the guidelines, a 15 to 30 percent margin difference.
Those two options, in app purchase and Stripe, are the whole picture for how money enters an app. But if your business model runs on subscriptions, there is a tool worth recommending on top of in app purchase, and it is the one thing we tell almost every subscription client to use: RevenueCat.
RevenueCat does not move money and does not replace either payment path. It sits on top of the store billing and handles the machinery around subscriptions: validating receipts, knowing whether a given user is currently entitled to premium, keeping that answer identical across iOS and Android, subscription analytics, and paywall testing. Building all of that yourself is weeks of unglamorous engineering that every subscription app needs and no user ever sees, and it does not change what the store charges either way.
It slots cleanly into either stack we build in, whether that is a Flutter app or a React Native one, so the recommendation holds regardless of the technology. If you have not settled that question yet, our guide to Flutter and React Native works through it. If your app has no subscriptions, you can skip RevenueCat entirely: it solves a problem you do not have.
Put together, the choice between the two options is mostly made for you by your business model:
That second line is worth pausing on, because it is the closest thing to free money in this article. A business selling real world services keeps the full payment and pays only its processor. If you are choosing what to build first and the two candidates are close, the services model starts with a structurally better margin than the content one.
What it costs instead is engineering around bookings, cancellations and payouts, which is scoped on booking app development.
Everything above is framed around Apple because Apple's rules are stricter and its review is where teams get caught. Google Play applies the same core test, digital goods go through its billing or a permitted alternative, real world services stay outside it, but the numbers changed materially in 2026 and are worth knowing before you model revenue.
From June 30, 2026, in the US, UK, and European Economic Area, Google separated its charge into two parts: a service fee that applies no matter how you collect payment, and a billing fee of 5 percent that only applies if you use Google's own billing. For auto renewing subscriptions the service fee is 10 percent at any revenue level, so most subscription apps using Google's billing still land at 15 percent, the same effective rate as before.
The difference is that you can now route payment through your own checkout or a web link and skip that 5 percent, paying your processor instead. For one time purchases, the service fee is 10 percent on your first 1 million dollars of annual earnings and higher above it, and Google is opening reduced rates for apps that join its new Apps Experience and Games Level Up programs.
Whether leaving Google's billing saves money depends entirely on transaction size. On a small monthly plan, a card processor's fixed per transaction fee eats most of the 5 percent you saved. On a large annual renewal, the processor's percentage comes in under 5 percent and you keep the difference.
So the honest answer is that Android's lower headline rate is real, but it only shows up in your accounts on higher value transactions, and it is a decision to make per product rather than a blanket win.
If your users are in the European Union, the numbers above are not the ones that will apply to you. On August 18, 2026 Apple announced new business terms for the EU, following its work with the European Commission, and they take effect on October 1, 2026. Every developer distributing apps in the EU moves onto one set of terms instead of the layered arrangement introduced after the Digital Markets Act.
What changes in practice:
The practical reading for a European business is that the arithmetic finally became possible to do in a meeting. Linking out at 15 percent is not automatically cheaper than in app purchase at 26 percent once you add your own processor, chargebacks and the engineering for a second checkout, and for a Small Business Program member paying 15 percent inside the app, a 10 percent link out saves only 5 points before those costs.
Model your own average transaction before choosing a path. We work through this with clients as part of scoping, and the wider European setup is on app development in Europe.
Since the April 2025 contempt ruling in the Epic litigation, apps on the US storefront may link out to a web checkout for digital purchases, and Apple currently charges nothing on those purchases. That zero is a court imposed position rather than a policy, and it is being decided right now.
On August 13, 2026 Apple filed a proposal asking the district court to approve commissions on external link purchases: 15 percent as standard, 10 percent for partner programs and subscription renewals, and 5 percent for Small Business Program members. Those rates are not in force. The district court has to decide what Apple may charge, Epic opposes any commission, and the Supreme Court has separately agreed to review the contempt finding itself.
Treat the current zero as a window, not a rate. If a web checkout is cheap for you to add and your audience is US heavy, the upside is real today.
If building it would consume a meaningful part of a first release, build the store path first, keep the checkout in the plan, and revisit when the court settles the number. Designing a whole business model around a rate that is still being argued in court is the expensive version of this decision.
30 percent as standard, and 15 percent if you are in the App Store Small Business Program or the subscriber has been paying for more than one year. In the European Union the standard rate becomes 26 percent from October 1, 2026. Apple takes nothing on physical goods or real world services.
For subscriptions in the US, UK and European Economic Area, 15 percent when you use Google's billing, made up of a 10 percent service fee and a 5 percent billing fee. Take payment through your own checkout and the billing fee falls away, leaving 10 percent plus whatever your processor charges. One time purchases carry a 10 percent service fee on your first 1 million dollars of annual earnings.
No. Payments for services delivered in the physical world are processed outside Apple's system, and Apple takes no commission on them. The 30 percent applies to digital content and features only.
Enroll in the App Store Small Business Program if your App Store proceeds were under 1 million dollars last calendar year. It takes minutes, needs no code change, and applies from approval onward rather than retroactively, which is why it belongs on the pre launch checklist.
Not inside the app on most storefronts. The US storefront currently allows linking out to a web checkout with no Apple commission while the court decides on a rate, the EU allows alternative processing and link outs at the rates above, and subscriptions sold on your own website have always been yours to process with Stripe.
No. Apple's StoreKit is enough on its own. RevenueCat replaces the custom backend work around subscriptions, which is why it earns its place on subscription products and adds little on apps without them.
Not directly, but they change which model is worth building. A services product that keeps the full payment and a subscription product that gives up 15 to 30 percent of it need different revenue assumptions before you commit a budget. What drives the build cost itself is set out in what an app development budget actually buys.
It rarely comes down to fees. Launch where your users are. The fee difference only becomes a real number at scale or on high value transactions, so for a first version it is not the deciding factor. What matters more is that your payment model is clean on whichever store you submit to first, because that is where a rejection would cost you a launch date.
Co-Founder and CEO, Apps Value
Thomas sets the company's long term strategy and direction, identifies market opportunities, and owns business development and client acquisition. He builds strategic partnerships that last beyond a single project. He works with clients from the first strategy conversation, so their business goals, not just their feature list, drive every decision.
Tell us what your app will charge money for and where your users are. We will tell you which payments run through the stores, which go through Stripe, what the current rates mean for your margin, and whether RevenueCat belongs in the build.
Book an intro call