PRICES CHECKED SEP 2026
The floor
Calculator
The cut
No invoice
What mine costs
vs an agency
Spend less
Run the numbers →
$99 a year to Apple, $25 once to Google. Everybody quotes those two numbers and stops. Then the domain renews at triple what you paid, the database wants $25 a month to stop falling asleep, the AI subscriptions you actually built the thing with cost more per year than Apple does, the privacy policy you need before review will pass you is a subscription too, and the store takes 15% of every dollar before you see it.
This page is the whole bill, itemized, with the real 2026 prices and links to where they came from. Check the boxes for what your app actually needs and it will total itself.
Most cost breakdowns online are written by agencies quoting you for the build. This one is not about the build. It is about the standing charges — the things that keep an already-finished app alive on a shelf, whether anyone downloads it or not.
Nearly all of them renew. That is the part people miss. A one-time build cost is a decision you make once; a $99 renewal is a decision you make every year for as long as the app exists, and the year you forget, the app comes down.
Every line is a real, current, publicly listed price — sourced at the bottom of the page. Uncheck anything you can live without and watch what happens to the total. Most people can live without more than they think.
One-time fees and annual fees, all billed up front.
Everything above, plus twelve months of subscriptions.
That is $0 a month, forever, to keep the lights on.
Excludes the platform and payment cuts — those come out of revenue, not your pocket. See section 03.
The costs above are what you pay to exist. This is what you pay to earn. It does not show up on a credit card statement — it is simply money that never reaches you, which is why it is so easy to leave out of a plan.
If you sell anything that unlocks something inside the app — a subscription, a premium tier, coins — it must go through the store's own payment system, and the store keeps a cut.
The headline rate is 30%. But if you earned under $1M last year, both Apple's Small Business Program and Google's equivalent drop you to 15% — and you have to enroll. It is not automatic. On a $4.99/month subscription that is the difference between keeping $3.49 and keeping $4.24.
Selling on your own website instead — a course, a service, a physical thing — is dramatically cheaper. No monthly fee, no setup fee. Add 1.5% for international cards and 1% if a currency has to be converted.
The 30¢ is the sharp edge. On a $2 sale it is a 15% fee on its own. Small transactions are where flat fees quietly eat a business.
A 99¢ app on the App Store at the 15% rate returns about 84¢. To clear the $99 Apple fee alone you need roughly 118 downloads a year — and that is before hosting, before the domain, before anything else on this page.
Run that same math against your own idea before you build it, not after. It is the single most useful hour you can spend.
Add every number in the calculator together and, for most solo apps, it lands somewhere under $600 a year. That is a phone plan. It is not why apps die.
Apps die from the costs on the right — the ones that are paid in weeknights, in momentum, and in the slow discovery that shipping was the beginning of the work rather than the end of it.
If you would have been paid $60 an hour for that time, a 200-hour build is a $12,000 app — ninety-odd times the cash cost of everything else on this page combined. You do not have to write yourself a check to have spent it. Price it honestly at the start and the "is this worth it?" conversation gets a lot shorter.
Review is free and usually fast, but a rejection costs you days, not hours — and first submissions get rejected routinely for things nobody warns you about: a missing privacy policy URL, an account you cannot delete from inside the app, a sign-in method Apple requires you to offer alongside your own, screenshots at the wrong sizes. Budget two or three rounds for a first app and you will not be disappointed.
If you register a personal Google Play developer account, you must run a closed test with at least 12 testers for 14 continuous days before Google will let you publish to production. Twelve real humans with Google accounts, recruited and kept engaged for two weeks. It costs $0 and it is the reason plenty of Android launches slip a month.
Every autumn a new iOS lands and something you wrote stops behaving. Apple deprecates APIs; certificates expire; a dependency you never think about publishes a breaking change. An app that is "done" still needs several days a year of attention just to stay exactly as good as it already was. Nobody budgets for standing still.
The moment a stranger can download your app, someone will email you at 11pm because it will not sync. Refunds, App Store reviews you cannot reply to properly, bug reports with no reproduction steps. It is not expensive; it is relentless, and it never appears in a cost estimate.
Domain, Apple membership, policy generator, LLC annual report — four separate dates, four separate cards on file, four separate ways to lose the thing quietly. Put every renewal date in one calendar the week you launch. The cheapest hour on this entire page.
The most expensive app is the one that got to 80% and stopped: every dollar and every evening spent, nothing shipped, nothing learned from a real user. Scope so that finishing is the likeliest outcome. That is the whole argument behind how I work, and behind the Lab.
Not a model, not a typical case — the real invoice list. Before Us is my own app: a daily-quote app of Black American and Black Diaspora voices, built in React Native, on the App Store, with a marketing site alongside it.
I am publishing it because every cost breakdown online is written by someone selling you a build. This is what one finished, shipped, one-person app costs to keep alive for a year — including the line that turned out to dominate it, which is not the one I expected.
Non-negotiable. Renews every twelve months or the app comes down, reviews and all.
The marketing site Apple requires me to point a privacy policy and support URL at. .app is a Google-run TLD that mandates HTTPS and prices a little above .com.
Hosting for that site. The free tier would carry it comfortably; I pay for Pro because it covers commercial use and the rest of my work sits on the same account.
Pro, plus a little compute. The free tier pauses a project after a week of inactivity — fine while building, fatal the moment real people are using it.
Hand-written. Apple will not review an app without a policy URL, but nothing says you have to rent one. I wrote mine against what the app actually does, which took an afternoon and has cost nothing since.
The subscription I design and build against, at the $100 tier rather than the $20 one.
Kept alongside it. Each is better at something, which is how most people end up paying for both.
Due every April whether the business earned anything or not. I pay it, but FEZ Media would owe it with or without this app, so it sits outside the total rather than inside it. Subtract it entirely if you ship as an individual; several states charge nothing at all.
Apple, the domain, hosting and the database. The entire cost of Before Us existing in the world.
Claude and ChatGPT. Two subscriptions, no servers, nothing shipped to a user — and more than double the running cost.
Two thirds of my annual bill is not infrastructure. Five years ago this line did not exist at all.
That is the part worth sitting with. I expected Apple and the database to be the story; they are less than a third of it. The single largest line on my invoice list is the tool I write and design with: twelve times the Apple membership, and nearly double everything needed to actually run the app.
One honest caveat: both AI subscriptions do double duty across client work and this app. I am counting them in full because I could not tell you truthfully which half is which — but if you already pay for a Max-tier subscription for other work, the marginal cost of an app like mine drops to about $920.
All in, $2,122 is about $177 a month. At 15% commission on a $4.99 subscription I keep $4.24, so Before Us needs roughly 42 paying subscribers to cover its own year — 48 with the LLC report. Forty-two is a number you can picture, which is exactly why it is worth working out before you build rather than after.
One warning before the chart, because this is where most comparisons cheat. Everything above is what it costs to run an app. An agency quote is what it costs to build one. Putting $2,122 next to $150,000 and calling it a saving is dishonest — they are not measuring the same thing.
So here they are on the only basis that compares fairly: cash out of pocket to reach a shipped version one. And note the part nobody selling you a build mentions — the running costs on this page do not go away when you hire someone. You pay those either way. An agency builds the app; it does not pay your Apple fee in year three.
Your own time is deliberately not in those bars. Put it in and the first one moves the most: 300 hours at $60 an hour is $18,000 of your life, which lands it squarely on top of the offshore bar. That is the real trade, and it is why the table matters more than the chart.
Money is the least interesting column here. What actually separates these paths is who carries the knowledge, and what it costs you the day you want to change something.
| Measure | Build it yourself | Offshore team | US freelancer | US agency |
|---|---|---|---|---|
| CASH TO V1 | $400 – $2,400 | $10k – $30k | $30k – $60k | $60k – $250k |
| YOUR HOURS | 200 – 400, building | 40 – 80, managing | 40 – 80, managing | 20 – 40, in meetings |
| TIME TO SHIP | 6 weeks to 6 months, entirely down to you | 2 – 4 months, plus timezone drag | 2 – 4 months | 3 – 6 months, discovery included |
| CHANGING IT | Free and immediate. You open the file. | A new contract every time | Hourly, if they are still available | A change order and a queue |
| WHEN IT BREAKS | You fix it, at whatever hour | You wait for their morning | Depends entirely on the person | A support retainer, often 15–20% a year |
| WHAT YOU END UP WITH | A product and the ability to make the next one | A codebase and a dependency | A codebase and one phone number | A polished product and an invoice trail |
| THE REAL RISK | You run out of evenings before you finish | You cannot tell good work from bad until it is late | They take a full-time job mid-build | You spend the whole budget proving the idea |
I take client work, so read this with that in mind. Hiring is genuinely the right answer when the deadline is real, when the thing is complex enough that a wrong architecture costs more than the agency does, when it has to be accessible and compliant on day one, or when your time is simply worth more spent elsewhere. Those are good reasons and they are common.
The bad reason is using an agency to find out whether anyone wants the thing. That is a $60,000 answer to a question a rough prototype answers for a few hundred — and if the answer comes back no, the money is gone and you have learned it the most expensive way available.
Not advice to be cheap. Advice to spend in the right order — which is to say, spend nothing until a real person has used the thing and told you it is worth their time.
A website costs $22 a year and reaches everyone on any device today. The App Store costs $99 a year, takes 15%, and adds a review queue between you and every fix you ship. Validate on the web; move native when the demand is proven.
Vercel's hobby tier and Supabase's free tier will carry a genuine product a long way. Move up when a limit actually bites — a paused project, a real user complaining — and not one month before.
The first-year price is marketing. Compare the renewal price, because that is the one you will pay for the next decade — it is routinely double or triple the teaser rate.
Halving the store's cut from 30% to 15% takes a form and a few minutes, and it is not applied automatically. If you sell anything in-app and earned under $1M last year, this is the highest-value form you will ever fill in.
In Maryland that is $100 to open and $300 every April, forever, before the app has earned a cent. There are good reasons to incorporate — liability, contracts, a co-founder — but "I am about to publish an app" is not one of them on its own. Talk to an accountant, not a blog.
$20 a month feels like nothing and stacks like everything — two of them outrun your Apple fee, your domain and your database combined. Pay for one, pay annually if the discount is real, and only step up to a $100–200 tier when you are genuinely hitting limits rather than idly wondering.
Every feature you cut removes a service, a subscription, a permission dialog, and a reason for review to reject you. The cheapest app is a small one that works, and it is also the one most likely to actually get finished.
The Innovation Lab is six weeks of taking one idea from a sentence to a working prototype — including the unglamorous session where we work out what yours will actually cost to run and whether the math holds. Six seats.
Prices move. Check the source before you plan around a number here. Nothing on this page is legal, tax, or financial advice.