How much does it cost to build an app like Uber or Bolt?

A ride-hailing app like Uber or Bolt costs somewhere between $35,000 and $300,000 to build. A first version for one city sits at the bottom of that. A multi-city platform with all the extras sits at the top.

That's a wide range. Here's why, and how to work out where you land.

Why is it so expensive? It looks like one app

It is not one app. It is three products that have to talk to each other every second.

1. 01The rider app. Book, track, pay, rate.
2. 02The driver app. Go online, accept a job, follow the route, see earnings.
3. 03The admin dashboard. Where you approve drivers, set fares, watch trips and handle complaints.

Behind those sits the part nobody sees: the system that matches a rider to the nearest driver in real time, works out the fare and moves the money. That hidden part is the biggest cost in the build.

What do the published numbers say?

Estimates published this year put it roughly like this.

• A first version for one city, with rider app, driver app and admin: $35,000 to $70,000, and three to four months of work.
• A production platform with several cities and surge pricing: $120,000 to $200,000.
• Everything Uber has, at Uber's scale: millions. You do not need this.

Those are agency prices in dollars. You will find cheaper and you will find dearer. Cheaper usually means a template with your logo on it. I'll come back to that.

What moves the price?

Three things do most of the damage.

Live location. Every driver's phone reporting where it is every few seconds, and every rider watching a car move on a map. It has to work on a weak signal and not kill the battery.

Matching. Who gets the job? The nearest driver by road, not in a straight line. And what happens when he ignores it. This is logic, and logic takes time to get right.

Payments. Cards are the easy part. Cash is the hard part, because the driver has the money and you need your commission. In a lot of markets mobile money matters more than cards.

After those three: driver checks, how many cities, how many kinds of vehicle, and whether you want it in both app stores on day one.

Can I buy a ready-made one for a few thousand?

Yes. There are white-label ride apps you can rent or buy cheaply. For testing an idea in one town, that can be the right move.

Here's the part people hate hearing. The moment your business is different from Uber, the template fights you. A different vehicle. A different way of paying. Your own drivers on your own terms. You end up paying to bend something that was never built for you.

It's not the price - it's the fit.

Where do founders waste money?

Building everything at once. Surge pricing, loyalty points, scheduled rides, a wallet, a chat. None of that matters until riders can book and drivers can get paid.

The version that makes money is small: book a ride, see it coming, pay for it, rate it. On the other side: get the job, find the rider, get paid. Plus an admin screen where you approve who drives. Launch that. Add the rest when real riders ask for it.

What does it cost after launch?

People forget this one. Maps charge per use. Text messages for login codes cost money. The payment company takes its cut. Then there are servers, updates, and the app stores changing their rules every year.

Budget a monthly figure for running it and looking after it. An app nobody maintains gets slower, then breaks, then gets pulled.

So what should you do first?

Write down what makes your service different from Uber or Bolt in your city. That one page decides whether you rent a template or build your own.

Then get the scope written before anyone writes code. We build apps this way: the screens agreed first, a test version on your phone during the build, then the stores. If you want a straight number for yours, tell me what it has to do.