The cheapest way to build a mobile app is your own hours plus one store fee.

What you cannot avoid paying

Start with the part that has a published price on it. Apple's enrollment page states it plainly. The Apple Developer Program is 99 USD per membership year, listed in local currency during enrollment, and prices may vary by region. Nonprofits, accredited educational institutions and government entities that meet certain requirements can request a fee waiver. Everyone else pays, and pays again next year.

Google is cheaper and simpler. There is a one-time US$25 registration fee, prepaid cards are not accepted, and you must be at least 18. Identity verification can ask for a government ID and a credit card in your legal name. Google says that if that information turns out to be invalid, the registration fee is not refunded. Read that sentence twice if you were planning to register under something other than your own name.

So the floor is $25 for Android forever, $99 a year for iOS, and $124 for both in year one. Then add the computer. A hand-built iOS app means a Mac running Xcode. That is zero if you already own one, or the largest line on the whole list if you do not. A builder that compiles in the cloud is the only route that removes it.

Those are entry fees, not the whole bill. Store commission once you charge for anything, the renewal in year two, and the surprises that only appear at submission are the fees at the end of the process. They deserve a page of their own rather than a line here.

Apple Developer Program, Become a member: terms and payment

Four routes, and who each one is cheapest for

Low cost app development is not one route, it is four, and which is cheapest depends on the resource you are short of. You write it yourself. An AI builder writes it and hands you the project. You assemble it in a no-code tool. Or you pay a person. Cash climbs across that list and hours fall, almost step for step, and that trade is the entire decision.

Writing it yourself is cheapest in money and dearest in everything else. Nobody should pretend otherwise. A first React Native app needs sign-in, a list, a detail screen and a release build the store will accept. That is weeks of evenings for somebody who already codes, and a course first for somebody who does not. It is the right answer when your time genuinely has no market price and learning is part of what you want.

The paid middle is where most people land, and the honest framing is a rate rather than a price. A subscription you keep for two months to ship one app beats a hundred unpaid evenings at almost any wage. It is also more expensive than nothing at all, if those evenings were never going to be used. That is the real content of free vs paid app builders: not which has more features, but whether the thing you are short of is dollars or weeks.

Paying a person is the expensive route and it is sometimes still the correct one, because a fixed quote moves the risk of never finishing onto somebody else. The cheapest app in the world is the one that gets built. A free one that stalls in month three cost you the whole idea.

Where free stops being free

An affordable app builder is not the same as a free one, and the difference shows up at the exit. Designing, building and previewing is free almost everywhere, because that part costs the vendor least. The paywall tends to sit on the way out: exporting the code, removing the builder's badge, pointing your own domain at it, or publishing under your own store account. Find that line before you spend a weekend, and find it on the vendor's own pricing page. We are deliberately not quoting anybody's plan here, because free tiers move and a price in an article ages badly.

Then there is the store side, which no builder can make free on your behalf. Google's own getting started page lists the $25 fee, then adds two conditions for personal accounts. Accounts created after 13 November 2023 must meet testing requirements before the app can reach Google Play. New personal accounts must also verify access to an Android device through the Play Console mobile app.

The testing requirement is the one that turns a free weekend into a month. Google's testing requirements page says a closed test needs at least 12 testers, opted in continuously for the preceding 14 days, before you can apply for production access. Finding twelve people who will install a half-finished app and leave it installed is a harder ask than $25. The rule is labelled personal accounts only. An organization account is a different path, and that is a real reason to register as one if you already have the entity.

None of this makes the cheap route wrong. It makes free a word about the building and not about the launching, and the launching is the half people forget to budget for.

Google Play Console Help, Get started with Play Console: registration fee and testing requirements

The bill the cheap route sends in weeks

Price your hours or you have not priced the app. Put any figure on an hour of your own time, even ten dollars, and the do-it-yourself route stops being free before the end of the first evening. That is not an argument for spending money. It is an argument for knowing which currency you decided to spend.

The largest real saving is not a tool, it is scope. Half the cost of a first app is features nobody asked for. An account system for an app with one user. A settings screen with three toggles. A server for data that never leaves the phone. Starting local-first, with the data on the device and no backend at all, removes the only recurring bill before it starts. That is most of what mvp app development is actually about, and it beats switching tools by a wide margin.

And here is the uncomfortable answer, the one a page selling a builder would leave out. Sometimes the cheapest way to settle your question is not to build an app at all. A landing page and a form will tell you whether anyone wants the thing, for the price of a domain, in an afternoon. Build the app when you have a reason to believe somebody will open it twice.