March 4, 2026

I cloned a $150/year fasting app with one prompt. It's now on the App Store.

Paid mobile apps as a business model are quietly dying. Not because people stopped wanting software. Not because the App Store disappeared. But because the cost of building most apps just went to zero.

A week ago I proved this to myself.

The app

I use an intermittent fasting app. I won't name it for obvious reasons.

From the data I was able to gather, it has roughly 30 million monthly active users. The paid version costs around $50 for three months.

Functionally, the app is quite simple. It has some nice UIs though.

It tracks fasting windows and shows you a visual history of your discipline. The core mechanic is psychological. When you see a long graph of days where you resisted eating after 4pm, you become less likely to break the streak.

And it works.

I was using the free version, but the interstitial ads were more than annoying. So I briefly considered paying, like the last time I was on a diet.

Then another thought crossed my mind.

Could I just clone it?

The experiment

I took screenshots of the app and sent them to Claude Opus 4.6.

My prompt was simple:

Describe the app in detail based on these screenshots and produce instructions for Claude Code to recreate it.

Claude generated a full specification.

I added a few instructions:

That was basically it.

For context, I have been developing mobile apps for more than ten years. But honestly, a non-technical person could write prompts like this too.

I then pasted the specification into Claude Code Opus 4.6.

It generated the entire project.

I did not rewrite the code.

I did not redesign the architecture.

I literally submitted the first thing that came out.

Shipping it

I created:

Uploaded the build.

And waited.

Three days later, Apple approved it. I assume the review queue is filling up quickly these days.

The app is now live:

https://apps.apple.com/pl/app/infastic/id6759506430

Everything about it is AI generated:

A week later

A week after release I actually tried using it myself.

I vibecoded a few improvements.

And honestly?

It's good.

Not a toy project. Not a broken prototype.

Just... a working fasting tracker.

What just happened

The important part is not that I cloned a fasting app.

The important part is how trivial it was.

For the majority of paid apps, the real moat used to be development cost.

Someone had to:

That barrier is collapsing.

If a $50 subscription app with tens of millions of users can be cloned in a few prompts and submitted to the App Store in a few days, the economics change.

The coming wave

Most paid utility apps are vulnerable.

Especially ones that are:

Anything with a clear UI and limited backend logic can now be recreated quickly.

Not perfectly.

But good enough.

And good enough is all it takes when the alternative costs $200 a year.

What might survive

Large apps will probably survive if they rely on things that cannot be cloned instantly.

Examples:

User generated content

Communities, social graphs, and user data are harder to replicate.

Or apps that rely on proprietary datasets.

Or apps where brand matters.

But standalone paid utilities?

The clock is ticking.

An uncomfortable realization

I work in mobile development.

This is literally my job.

And I am the one writing this.

But after running this experiment I cannot escape the conclusion:

paid apps are already dead.

They just do not know it yet.