Quiet Mind Creative
How Much Does a Custom App Cost?
The price of a custom app comes out of a few decisions: accounts, integrations, a second platform, offline sync, and App Store review. Here is what each one adds.

You probably have a number in mind. Maybe it came from a friend’s project, another studio’s quote, or a no-code tool’s monthly price, and you’d like to know whether it’s in the right neighborhood before you book a call. Most pricing pages say “it depends,” which is true and doesn’t help.
What it depends on is a short list, and you can answer most of it before you talk to anyone. The cost of a custom app is an output of a few product decisions: whether people sign in, what it talks to, how many platforms it runs on, whether it syncs, and whether it goes through the App Store. For the work we do, most projects land between $5,000 and $40,000, and those decisions are what place an idea in that range.
Website or web app: where the price jumps
The biggest jump comes when a website becomes a web app. If people mostly read, compare, and get in touch, it’s a website. If they sign in, save things, and come back to get work done, it’s a web app, and the software is holding something on their behalf. Where that’s stored, who can see it, and how long it’s kept are now part of the job.
Two of our own products (every example here is ours, not client work) show how far apart the ends are. Meal Plans is a React and Vite app that keeps your plans in the browser’s local storage. There’s no account and no server, and if you want a plan on another device, you export it as JSON and import it there. Burn After, our app for sharing files through single-use links, is the other end. Its interface is a form and a link, and every layer underneath is work a simpler app doesn’t need.
What sits under a single upload
Here’s what happens when someone shares a file through Burn After:
- The file is encrypted in the browser with AES-GCM before upload, and the key travels after the
#in the link, so it never reaches a server. That’s cryptography code plus a clear interface state for every step, because a burned link can’t be brought back. - Senders sign in, and their requests go through API Gateway to eight Lambda functions. Recipients just get a link.
- Encrypted files go to S3 and records to DynamoDB, each with rules for how long it lives.
- A cleanup job scheduled with EventBridge runs every minute and deletes whatever has expired. Nobody watches it run, so it needs testing as careful as the parts people see.
- Everything is defined in code with the AWS CDK, which takes longer up front and turns standing up a new environment into one command.
None of this is expensive to run at small scale. At AWS’s published US East prices, API Gateway charges $3.50 per million REST API calls, Lambda $0.20 per million requests plus compute time, and DynamoDB $0.625 per million writes on demand. Unless you expect serious traffic from day one, the hosting bill probably won’t decide your budget. The hours will: designing each layer, building it, and making the seams between layers hold.
Scope can move after launch, too. Burn After’s encryption moved into the browser after its Product Hunt launch, because the feedback was about trust. A good budget leaves room for what real users tell you.
Five decisions that move the number
Do people sign in?
Accounts look like one feature and turn out to be several: sign-up, sign-in, password resets, expiring sessions, permissions, and a database of people you’re now responsible for protecting. Apple adds rules of its own for iPhone apps. Under the App Store Review Guidelines, if people can create an account, they have to be able to delete it inside the app (5.1.1(v)). A third-party login like Google generally has to sit beside an equivalent option that limits data collection to name and email (4.8). And App Review needs a way in, usually a working demo account (2.1).
The cheapest account system is the one you don’t build, so the useful question is who actually needs one. Burn After puts accounts on one side only. Momentum and Tango have none at all.
What does it talk to?
Every integration brings someone else’s system into yours: their authentication, their limits, their outages, and their changes on their schedule. Each one needs an error state for when it’s down. Reading from a service is the cheaper kind. Two-way sync costs more, because someone has to decide what happens when both sides change. Tango syncs habit completion with Apple Reminders in both directions, and its design settles that question up front: the reminder stays the source of truth for scheduling, and Tango tracks completion.
So count them. Payments, email delivery, a CRM, a calendar, a sign-in provider: each one comes with its own setup, failure modes, and upkeep.
How many platforms?
Closely related platforms can share a lot. SwiftUI makes sharing code between iPhone and iPad practical (Rotate Meals runs on both), though a good iPad app still deserves its own layout decisions. Across different technologies, like iPhone and Android, or a native app and a web app, the second platform is closer to a second product, with its own code, testing, release process, and yearly round of operating system changes. The cheaper path is usually one platform first, chosen by where your users already are, and the second once the first has proved the idea.
Does it work offline, and does it sync?
Working offline on one device is cheap when the data lives on the device anyway. Rotate Meals keeps meals, plans, and ingredients on the phone, with no account. Syncing across devices is where the cost shows up, because two devices can change the same thing while one is offline, and something has to reconcile them. Building that yourself means a server, accounts, and rules for conflicts.
Momentum and Tango take a different route. Their data is stored on the device with SwiftData and synced through the person’s private iCloud database with CloudKit, so there’s nothing to sign up for and no user database to protect. That one choice removes a large share of what a web app like Burn After needs, and it has costs of its own. It’s built for people on Apple devices, so an Android version later would need its own answer for sync. It also makes early decisions stick: Apple’s documentation says you can’t rename or delete CloudKit record types and fields in production, so the data model deserves more thought before launch than it would with a database you control.
Does it go through the App Store?
The fee is the small part: the Apple Developer Program costs $99 a year. The work is the listing, the screenshots, the privacy disclosures, and a privacy policy linked from both the listing and the app (5.1.1(i)); Momentum’s lives on this site. Every update goes through review, so each release has a step you don’t control. Selling something digital inside the app, like Momentum’s one-time Pro unlock, generally means Apple’s in-app purchase (3.1.1), and Apple takes a commission: 15 percent for developers in its Small Business Program. And the app has to be more than a repackaged website (4.2), which rules out the cheapest way to “have an app.”
Rough ranges
Here’s how those decisions add up, with our own apps as reference points for each shape. Each range covers design, development, and launch, including App Store submission for native apps:
- Runs in the browser, data stays on the device. Interface, logic, local storage, and export, like Meal Plans: $5,000 to $11,000.
- A native iPhone app with iCloud sync, through the App Store. No accounts and no server, like Momentum or Tango: $11,500 to $24,000.
- A web app with accounts and a backend. Sign-in, an API, a database, file storage, scheduled jobs, and infrastructure in code, like Burn After: $20,000 to $40,000.
On top of any of these:
- A second platform adds 15 to 30 percent when it’s closely related, like an iPad version of an iPhone app, and 50 to 80 percent when it’s a different technology.
- Each integration adds $1,000 to $3,500 for a one-way connection, like reading from a service or a one-time in-app purchase, and $3,000 to $8,500 for two-way sync.
Running costs are separate: hosting and third-party services, Apple’s $99 a year, and maintenance, which for us typically runs $500 to $1,750 a month.
What the cheap option costs later
A low quote is usually low because something isn’t in it. Sometimes that’s exactly right, since a smaller first version that works beats a bigger one that almost does. The trouble comes when what’s missing is invisible at launch: the error and empty states, infrastructure that lives in someone’s memory instead of in code, the account deletion App Review will ask for, tests for the job that runs at 3 a.m.
Then there’s the year after launch, whether or not anyone touches the code. Chrome and Firefox ship new versions every four weeks. Apple previews the next iOS in June and ships it in September. Security fixes land in the libraries your app depends on whenever they’re found. If a quote doesn’t say who handles that, the answer is probably you. It’s the work our support and maintenance covers.
No-code tools deserve a fair hearing here. For an internal tool or a first test of an idea, one can be the cheapest route that works. Before you commit, check what the plan charges as users and usage grow, and what you can take with you if you outgrow it.
How to compare quotes
Before you contact anyone, write down your answers to the five questions: who signs in, what it talks to, which platforms, whether it syncs, and whether it goes through the App Store. Send them with your first message, and ask each studio, freelancer, or tool what its price assumes for each one, and who looks after the app once it’s live. Two quotes that are far apart are often pricing different apps, and this is how you find out. If you’d like a second opinion on which of these your idea really needs, tell us about it.
How we can help
Web app design and development
Software people sign in to and come back to, from the interface down to the infrastructure.
Swift and SwiftUI apps that behave like they belong on the device, shipped through the App Store.
Ongoing support and maintenance
Updates, fixes, platform changes, and steady improvements for websites, web apps, and native apps, long after launch.

