How much does an MVP cost? What actually moves the price, with worked examples at sample rates

So what's the short answer?

A first version of an app or web product, built by a small senior team, usually lands somewhere around $15,000–60,000 at our sample rates. That's a wide range on purpose. A booking tool on the web with sign-in and an admin page sits near the bottom of it. A marketplace on iOS and Android with payments and chat sits towards the top, and can go past it.

Every figure in this article is a sample, worked out with the same model as our cost estimator, so you can check the arithmetic. A real quote comes after a short scope, once someone has looked at what you actually need.

What counts as an MVP, for pricing?

Here, an MVP is the smallest version of your product that a real customer can use for the job it exists to do, built well enough that you'd put your name on it. Not a clickable mock-up, and not a demo that only works on the founder's laptop: real sign-in, real data, released to an app store or the web.

The definition matters because most of the wild price differences people hear about come from comparing different things. A prototype in a design tool might take a week. A production first version with payments takes weeks to months. Both get called an MVP. When you compare quotes, check that they're pricing the same thing.

What actually moves the price?

Nearly all of the cost is people's time: design, engineering, testing and release. Anything that adds days adds cost. Five things add the most days.

Platforms: iOS, Android, web, or a mix

Each platform after the first adds work, even with one shared codebase. Screens behave differently, each store has its own review rules, and everything has to be tested on each. In the estimator's sample model, every extra platform adds about 40% to the build rather than doubling it, because iOS and Android share one React Native codebase. Two fully separate native apps cost more than that.

The usual move is to launch where your first customers already are, and add the next platform once the idea has proved itself.

Features, and which ones are heavy

Features aren't equal. Some are a few days of well-trodden work; others hide weeks of edge cases. In the sample model, accounts and login take about 5 days, chat about 10 and computer vision about 14. Heavy features share a trait: they involve real-time data, money, other people's systems, or a model that has to be trained and checked.

• Usually light: sign-in, notifications, search and filters, analytics, a second language.
• Usually heavy: payments and subscriptions, chat, offline sync, an admin back office, bookings with reminders and no-shows, anything with AI or computer vision.

Every feature you can move to version two comes straight off the price. It's the biggest lever you have.

Design: clean, custom or signature

A clean design built from a well-made component library is quicker than a fully custom look, and custom is quicker than one with signature motion and illustration. The gap is smaller than people expect (the sample model puts clean at about 12% less than custom, and signature at about 20% more), but it touches every screen.

Where you're starting from

Starting from an idea means time deciding what to build before design begins; the sample model allows 5 days for it. Finished designs shorten that. A live product adds a different cost: time to read an existing codebase and release into it without breaking what customers already use.

Who builds it

Freelancers, agencies and small senior studios charge different day rates and work at different speeds. A lower rate isn't always a lower bill: rework, handovers between people and missed edge cases all cost days. When you compare quotes, compare what's included as well as the total: design, testing on real devices, release, fixes after launch, and whether the code ends up in your own accounts.

What might a few typical projects cost?

Three made-up projects, priced with the estimator's sample model. They show how the levers add up; they aren't past clients and they aren't quotes. Open any of them in the estimator and change one thing to see what it does to the number.

Look at what moved each one. The marketplace costs more because of a second platform and several heavy features (payments, chat and an admin back office), not because it's "an app". The AI assistant is a small feature list, but working inside a live product adds days of its own.

How can I lower the cost without cutting corners?

• Cut features, not quality. Postpone anything a first customer could live without for three months.
• Launch on one platform first, the one your first customers use.
• Use a clean, component-based design for version one, and invest in a custom look once you know what people use.
• Bring what you have: designs, a written brief, products you like and dislike. It shortens the deciding phase.
• Ask for a fixed-fee scope before a build. 2 weeks of scoping costs less than six weeks building the wrong thing.

What not to cut: testing on real devices, crash and error reporting, and a sensible release process. They're cheap to do now and expensive to add after launch.

What costs come after launch?

The build isn't the whole bill. Plan for these from the start, so they don't arrive as surprises:

• Hosting and third-party services: often small at first, growing with your users.
• App store developer fees, and payment processing fees on every sale.
• Fixes and small changes once real people start using it in ways nobody predicted.
• Keeping libraries, frameworks and operating-system support up to date, a little every few months.

Our builds include 30 days of fixes after launch. After that, some teams keep a small team on a monthly plan and some take everything in-house with documentation and a handover.

How do I get a number for my project?

It depends how far along you are. For a rough range in a minute, use the estimator: circle the features, pick platforms and design, and it shows a range and a timeline at sample rates, with a one-page PDF you can keep. For a real number, send a short brief. If it's a fit, a 2 weeks scope sprint ends with a clickable prototype and a fixed quote for the build.

Not sure what to put in a brief? Our guide to writing a software brief a studio can quote has a checklist and a template.