How much does it cost to build an app? A 2026 breakdown for founders

1. The short answer

In 2026, a working first version of a web app — one core flow, real users, payments if you need them — lands between $0 and $60,000 depending on how you build it. The distribution looks roughly like this:

Approach	Typical range	Timeline
No-code build it yourself	$0 – $5,000	1–4 weeks
No-code with a consultant	$3,000 – $15,000	3–6 weeks
Freelance developer, custom build	$8,000 – $30,000	6–12 weeks
Small studio, custom build	$20,000 – $60,000	4–8 weeks
Agency, custom build	$60,000 – $150,000+	3–6 months
Full product with native iOS and Android	$150,000 – $400,000+	6–12 months

Notice that the cheapest and most expensive rows differ far more in coordination overhead than in lines of code. Two people who ship every week cost less to run than an organisation that meets about shipping.

2. What you're actually paying for

When a builder quotes you, the number is made of four things. Ask which is driving yours.

Design and product decisions

Deciding what the screens are, what order they come in, and what happens when someone does the unexpected thing. This is where good builders earn their fee — a bad decision here can cost weeks later.

Engineering hours

The visible part. A single core flow in a modern stack is typically 80–200 hours of senior work. That number changes surprisingly little between ideas; it changes a lot with the number of edge cases you insist on covering.

Integration risk

Payments, email, calendars, mapping, e-signature, accounting. Each is a day or two to wire up and a permanent source of support requests. Five integrations can add 20–30% to a build.

Accountability

Someone who is on the hook for launch day, not just for delivering files. You are paying for reduced risk. Agencies charge for it explicitly; freelancers usually don't, which is why their quotes are lower — and why so many projects stall at 85%.

3. Prices by stage

Stage 1 — prove the idea exists ($0 – $5,000)

A landing page, a form, a payment link, and a manual process behind it. If you can't sell ten of these, the software was never the problem. Time: a weekend to two weeks. This stage is almost always the correct starting point, and almost always skipped.

Stage 2 — first working version ($20,000 – $60,000)

Real accounts, a real database, one core flow that works end to end without you, and payments. This is what "MVP" should mean. Time: four to eight weeks with a small team.

Stage 3 — product with a team around it ($80,000 – $250,000)

Multiple user types, an admin surface for support, automation of the manual parts, notifications, and integrations your paying customers already use. This is a year of work, not a quarter, and it is where most app-companies actually begin.

Stage 4 — scale and mobile apps ($150,000+)

Native apps, infrastructure that survives traffic, compliance work, and a small engineering team. Nothing at this stage is optional for the business it serves, which is why the cost is what it is.

4. The five things that raise the bill

• A native mobile app. Two codebases (iOS and Android), app store reviews, device testing. Typically 2–3× the cost of the same product as a responsive web app. Justified when you need the camera, push, offline use, or a home-screen habit.
• Many user types. Each role multiplies permissions, empty states, edge cases, and testing. Three roles is roughly double the work of one.
• Real money and real legal exposure. Regulated industries (health, finance, insurance, anything carrying personal data) mean extra controls, extra review, and extra time. This is real cost, not padding.
• Third-party systems you don't control. Old CRMs, accounting packages, hospital systems, government portals. The cost is not the integration itself but the waiting and the undocumented behaviour.
• Changing your mind mid-build. A pivot at week five costs more than the same pivot at week one. Fixed-price builds usually price this risk in; hourly builds simply bill you for it.

5. The costs nobody quotes you

The build is usually the smaller half of the first year. Budget for these separately, or they'll eat the build budget.

Item	Typical cost
Hosting and infrastructure (early stage)	$0 – $60 / month
Email and transactional sending	$0 – $40 / month
Domains, certificates, fonts, stock assets	$50 – $300 / year
Payment processing	~3% of each transaction
Ongoing fixes and small changes	$500 – $2,000 / month
Customer acquisition (the big one)	Often equal to the build budget over year one

Founders routinely spend $40,000 on software and $0 on getting it into people's hands, then conclude the software didn't work. Marketing isn't optional; it's the other half of the product experiment.

6. How to get more for the same money

• Buy a narrower first version. Halving the scope cuts cost by more than half, because fewer features mean fewer interactions between features.
• Do the awkward work by hand at first. Approvals, matching, onboarding calls, even fulfilment. Manual behind the curtain costs a few hours a week and saves weeks of build.
• Reuse before you invent. Unless your entire value is a novel interface, a well-chosen template saves money you can spend on the part that's actually yours.
• Own your accounts. Hosting, database, payment processor, and domain should be in your name from day one. It costs nothing and makes switching builders survivable.
• Fixed price, clearly defined. Not because it's always cheaper, but because you can't budget against a number that moves.

7. What we'd budget in your position

If you're pre-revenue with an idea and a modest budget: $0 to $2,000 for stage one, spent on a page, a payment link, and thirty conversations. Come back when you have evidence.

If you have ten paying customers and a manual process falling over: $20,000–$50,000 for a first working version, plus roughly 10% of that per month afterwards for care and changes.

If you already run a business with real revenue and you're losing money monthly to a broken process: the calculation is different — build the thing that stops the leak, and measure it in months to payback, not in months to launch.

Related guides

For what a first version should actually contain, read MVP development for startups. For the build-it-yourself comparison, read how to build an app with no coding experience.

And if you have a quote in hand and want a second read on whether it's fair, send it across — we'll tell you plainly, even when we're not the cheaper option.