What does it actually cost to start an app in 2026?

In money, less than most people expect: you can run a real product with paying customers for roughly
**$5 to $30 a month** in infrastructure, plus payment processing of about 2.9% plus 30 cents per
transaction. In time, more than most people expect: **four to eight weeks** of building before the
first customer can do anything useful, and almost all of it goes to work that is identical in every
product. The time is the real cost, and it is the one worth attacking.

Here is the breakdown behind those numbers.

## The money, itemized

### Before you have customers

| Item | Realistic cost | Notes |
|---|---|---|
| Domain | $10 to $15 per year | The one unavoidable purchase |
| Server | $5 to $20 per month | A small instance handles far more than people assume |
| Database | $0 | SQLite on the same box is genuinely fine at this stage |
| Background jobs | $0 | If your queue runs on your database, there is nothing extra to pay for |
| TLS certificate | $0 | Automatic |
| Email sending | $0 to start | Most providers have a free tier in the thousands of messages |
| Error tracking | $0 to start | Free tiers are generous |
| Payment provider | $0 | No monthly fee, they take a cut of transactions |
| **Monthly total** | **$5 to $30** | |

That is not a trick. A single small server running one application with a local database will serve a
surprising number of customers before anything needs to change.

### The costs people forget

- **Apple Developer Program: $99 per year**, required to put anything in the App Store. Google Play
  is a one-time $25.
- **Both stores take a cut** of in-app purchases, currently up to 30% and lower for smaller
  developers. If you sell a subscription inside a mobile app, this is a much bigger number than your
  server bill. See [getting your web app into the app stores](/blog/your-app-in-the-app-stores).
- **A business entity**, if you need one, varies enormously by country.

### What gets expensive later, and what does not

The instinct is to over-provision for scale that has not arrived. Resist it. The costs that actually
grow are:

- **Email volume**, once you are past a free tier.
- **Anything billed per seat** across your team.
- **Managed everything.** A managed database, a managed queue, a managed cache, and a managed search
  index will quietly turn $10 a month into $200 a month while serving the same number of customers.

That last one is where most early budgets go, and it buys very little at a stage where you have one
server and one developer.

## The time, which is the actual cost

Here is where four to eight weeks goes for a competent developer building from an empty directory.
Treat these as ranges, not promises, because they vary with experience and how much you cut.

| Work | Typical time |
|---|---|
| Accounts, sign-in, sessions, social login | 5 to 10 days |
| Payments, webhooks, subscription state, entitlement | 5 to 10 days |
| Email: provider, templates, DNS records, deliverability | 2 to 4 days |
| Separating customer data, permissions | 3 to 6 days |
| Background jobs | 1 to 3 days |
| Security hardening, rate limits, data export and deletion | 3 to 5 days |
| Deployment, TLS, CI | 2 to 5 days |
| **Subtotal, before any of your product** | **21 to 43 days** |
| Your actual product | however long it takes |

Put a cost on that. At a contractor rate of $600 a day, the subtotal alone is $12,000 to $26,000. If
you are doing it yourself, it is four to eight weeks you are not spending on the thing that
differentiates you, and four to eight weeks is long enough for the idea to lose its urgency.

That is the real cost of starting an app, and it has almost nothing to do with your server bill.
Those ranges are estimated from the outside. There is a measured version of this same table, counted
line by line against a real codebase, in [how long it takes to build the parts every app
needs](/blog/how-long-it-takes-to-build).

## Does a coding agent change this?

Partly, and less than the marketing suggests.

An agent will absolutely produce all of the above faster than you would type it. What it will not do
is make the review cheaper. You still have to check the four places where [generated code looks
right and is not](/blog/what-your-agent-should-not-build), and those are the slowest things to
review because there is no failing test to point at.

The other thing that happens is that [most of the first run disappears into
plumbing](/blog/where-the-first-run-goes) rather than into your idea, because the plumbing is what
has to exist first.

So an agent compresses the twenty-one to forty-three days meaningfully. It does not remove them, and
what is left is the part you least want to be responsible for.

## The way to actually cut it

Do not build the identical part. Start from a codebase where it already exists, is tested, and is
documented, and spend your weeks on the part that is yours.

That is what One Shot is, and it costs $299 a year at the regular price. Against a subtotal of
$12,000 to $26,000 of equivalent work, that arithmetic is not subtle. The honest caveat is the one
worth repeating: it is a Ruby on Rails codebase, and if nobody on your team can maintain that, none
of this arithmetic applies to you.

There is a full list of what that includes, and what it deliberately does not, in [everything One
Shot gives you on day one](/blog/whats-in-the-box).

## The summary

- **Infrastructure is cheap**: $5 to $30 a month is a real number, not a starter tier you will grow
  out of next week.
- **Mobile is the expensive channel**, mostly because of store fees, not because of hosting.
- **Managed services are where budgets quietly go**, and at your stage they buy very little.
- **Time is the real cost**, it is four to eight weeks, and roughly all of it is work that is the same
  in every product.

Attack the last line. The others largely take care of themselves.