Which vibecoded apps make money today?

Are vibecoded apps actually making real money today?

Yes. Vibecoded apps are making real money today, with credible public cases ranging from roughly $12,000 a month to $3 million in annual recurring revenue.

That already settles the easiest part of the question. We found consumer apps, creator tools and vertical SaaS products that were built substantially through tools such as Cursor, Claude Code and Lovable and then turned into paying businesses.

The harder part is figuring out which examples still deserve to be taken seriously. Vibe coding produces unusually noisy revenue claims. A founder can launch on Monday, have a viral week, multiply the result by twelve and announce a seven-figure ARR business before anyone knows whether customers will stay.

Once we separate actual collected revenue from annualized launch spikes, recent numbers from old screenshots, and genuinely vibecoded products from normal startups that happen to use Cursor, the list gets much shorter. It also gets more convincing.

What should actually count as a vibecoded app?

For this analysis, a vibecoded app needs to have been built substantially by directing AI to produce the software, rather than by a conventional engineering team that happens to use AI coding tools.

The distinction has become more important lately because Cursor, Claude Code and similar tools are now normal parts of many developers' workflows. If every startup whose engineers use Cursor counts as vibecoded, the term tells us almost nothing.

Payout sits clearly inside our definition. RevenueCat says the entire app, including its design, code and assets, was produced through Claude Code and Cursor without a line of code being written manually. Stanley also qualifies: Stan's founders described building the original product during a 14-day AI-heavy coding sprint. Products created from scratch in Lovable, including Lumoo, ShiftNex, Klar, Plinq and QuickTables, fit naturally too.

We therefore exclude companies such as Lovable and Cursor themselves. They make money by selling vibe-coding tools. The question here is whether the applications created with those tools can themselves become businesses.

Can we trust the revenue numbers people post about vibecoded apps?

Some vibecoding revenue claims are unusually solid, while others should be read as evidence of early demand rather than proof of a durable business.

Payout gives us one of the cleanest early measurements. During RevenueCat's Shipaton, RevenueCat recorded $30,017 in revenue, 1,750 paying subscribers and more than 17,000 users. Those are actual payments observed through the subscription infrastructure rather than a founder multiplying one good day by 365.

Stanley sits one level below that in verification. Its founders have repeatedly disclosed the product's revenue, including through Business Insider, and the numbers have progressed over time rather than appearing once in a launch tweet. They are still company-reported figures rather than audited accounts.

QuickTables gives us a different check. Its founders said the restaurant software had passed €100,000 in annualized revenue, and the company was later acquired. We do not know the purchase price, so the acquisition cannot validate the precise ARR figure, but an actual buyer adds more weight than a screenshot on X.

The weakest figures are usually very young ARR claims. If an app gets $10,000 of recurring revenue in its first month, calling that "$120,000 ARR" is mathematically normal. We still have no idea whether month two will be $15,000, $10,000 or $2,000. Throughout this article, we keep that distinction visible.

Type of evidence	How much weight we give it	Example
Payment-platform or transaction data	High	Payout
Repeated founder disclosure through established reporting	Fairly high, but self-reported	Stanley
Revenue claim followed by an acquisition	Useful additional validation	QuickTables
Founder case study published by the building platform	Useful, with obvious selection bias	Lumoo, ShiftNex, Klar, Plinq
ARR extrapolated immediately after a viral launch	Evidence of demand, weak evidence of durability	Early fly.pieter.com

How common is it for Lovable apps to make money?

Making money with Lovable is clearly happening at meaningful scale, but most Lovable builders still have no product revenue.

Lovable's Build Economy study surveyed more than 14,300 users. Some 10.7% said they were making money directly from their product, while another 9.1% made money through a mix of their product and client work. Together, 19.8% reported at least some direct product monetization.

Another 15.3% earned through consulting or client work. Meanwhile, 60.5% said they had not monetized yet but planned to.

The denominator has become huge. Lovable currently says more than 60 million projects have been created on the platform, with around 1.2 million new projects being built each week. Those 60 million projects obviously do not represent 60 million serious startups. They include prototypes, personal tools, internal apps, landing pages and abandoned experiments.

So the tempting "one in five Lovable apps makes money" line is too loose. The survey measures builders, not randomly selected projects. Still, monetization has plainly moved beyond a handful of famous screenshots: thousands of people in a large user sample say they are already earning from products they built.

Which vibecoded apps are making serious money right now?

The strongest current list includes Stanley, Payout and Stoppr, while several Lovable-built businesses have substantial but less frequently updated public revenue figures.

Freshness is important here. We would rather show an older number honestly than quietly present it as today's revenue. For that reason, the table separates recent operating disclosures from the latest public figure we could find.

App	What it sells	Latest useful revenue evidence	How we treat it today
Stanley	AI content system for creators	About $3M ARR across its LinkedIn and Instagram products	Recent, active and repeatedly disclosed
Payout	Class-action settlement discovery	$80K MRR after four months	Recent RevenueCat disclosure
Stoppr	Consumer app for quitting sugar	About $12K/month after five months	Recent founder disclosure through Starter Story
ShiftNex	Healthcare workforce software	$1M ARR within five months	Strong founder/platform claim, less fresh
Lumoo	AI content software for fashion brands	$800K ARR within nine months	Strong founder/platform claim, less fresh
Klar	AI learning app	€130K ARR in its first 30 days; Lovable now also highlights a €7.5M ARR partnership announcement	Early revenue verified only through company/platform reporting
Plinq	Public-record and safety app	R$2.2M ARR after roughly three months	Strong early claim, less fresh
QuickTables	Restaurant software	More than $100K annualized before acquisition	Real commercial outcome, no longer an independent startup
fly.pieter.com	Browser flight simulator	Hit a $1M annualized run rate shortly after launch	Historical hit; the founder now classifies it as "made money but not sustainable"

Is Payout still the cleanest vibe-coding success story?

Payout is currently the cleanest example we found of a fully vibecoded consumer app becoming a meaningful subscription business.

RevenueCat recently used Payout founder Connor Burd as a growth case study and said the app went from zero to roughly $80,000 MRR in four months. That update is especially useful because RevenueCat had already observed actual payments from Payout during Shipaton, so we are not relying on a single founder screenshot.

Payout helps people find class-action settlements they may qualify for. The proposition is easy to understand: the customer may get money. That gives the app a much clearer consumer payoff than another generic productivity tool.

The growth story is also less magical than the "AI built the whole thing" headline suggests. Burd says Payout grew through influencer partnerships and then increasingly through paid acquisition. RevenueCat's recent discussion with him focused heavily on how he finds creators, tests campaigns and moves winning creative into Meta ads.

The code was produced unusually fast. The customer-acquisition system still had to be built the old-fashioned way: test, measure, lose money on some experiments and scale the ones that work.

Did Stanley really become a multi-million-dollar vibecoded product?

As seen above, Stanley is currently the clearest multi-million-dollar vibecoded product we found, with Stan reporting about $3 million ARR from the product line.

Stan cofounders John Hu and Vitalii Dodonov built the original LinkedIn product during a 14-day sprint. The first version reached roughly $200,000 ARR within six weeks and crossed $1 million ARR a few months later. Stan then launched an Instagram version and expanded the broader Stanley product.

More recent disclosures from Hu put the two Stanley products at about $3 million ARR combined. Stan itself is much larger, at roughly $41 million ARR, so Stanley still represents a minority of the parent company's revenue.

That context makes the example more interesting. Stanley did not start from zero distribution. Stan already had a large base of creators, years of customer knowledge, an established brand and working billing infrastructure. Hu and Dodonov could spend two weeks building because they already knew whom they wanted to sell to.

They also did user interviews before automating everything. Hu has described manually acting like "Stanley behind the curtain", producing ideas for potential customers and seeing whether people valued the output. By the time the software appeared, part of the product question had already been answered.

The 14-day sprint gets the headline, but the real advantage was the combination of fast building, an existing audience and unusually clear customer knowledge. That is much harder to copy.

Can a complete beginner really vibe-code a $10K/month app?

As pointed out above, Stoppr is already around $12,000 a month even though its founder came into the project without a software-development background.

David Attias had spent years working in finance before building Stoppr, a mobile subscription app designed mainly to help younger women reduce sugar consumption. According to a recent Starter Story breakdown, the product reached roughly $12,000 a month around five months after launch.

The origin story is refreshingly unromantic. Attias saw another consumer app working extremely well, studied its onboarding and business model, then adapted the approach to another habit category. He used AI coding tools to build the product rather than learning traditional mobile development first.

Nothing here required inventing new technology. Stoppr took a proven consumer-app structure, changed the problem being solved and executed quickly enough to test whether people would pay.

For a first-time vibe coder, that route looks far more reproducible than trying to invent the next software category from scratch.

Are companies paying real B2B money for Lovable-built software?

Yes. Some of the biggest disclosed Lovable-built businesses are selling ordinary B2B software into industries such as healthcare, fashion and restaurants.

Lovable's latest broad update on businesses built with its platform still highlights ShiftNex at $1 million ARR within five months. ShiftNex sells healthcare workforce software and had more than 5,000 healthcare users when the figure was published.

Lumoo reached roughly $800,000 ARR within nine months by selling AI-assisted fashion content and merchandising software. Its customer list has included established Nordic fashion businesses, and the product covers workflows such as visualizing products before physical samples exist, creating campaign assets and adapting content across markets.

QuickTables went after restaurants with a far less fashionable product: websites, direct ordering, SMS marketing and loyalty software. It crossed an annualized six-figure revenue level before being acquired.

The three companies have very little in common technologically. Commercially, the pattern is obvious. Healthcare staffing, fashion production and restaurant operations already have budgets attached to them. A founder only needs a relatively small number of business customers before the revenue becomes meaningful.

This is why vertical SaaS looks particularly well suited to vibe coding these days. AI can make the software much cheaper to create, while knowledge of an awkward industry workflow remains valuable.

Can vibecoded consumer apps make money without a huge existing audience?

Yes. Klar, Plinq and Stoppr show that a vibecoded consumer app can make money without a famous founder, although none of them relied on passive app-store discovery.

Klar is a good example of how quickly the distribution strategy can become more important than the build itself. Three founders created the AI learning product with Lovable and reported €130,000 ARR in the first 30 days. Lovable now says the product serves more than 6,000 students and highlights a partnership announcement carrying up to €7.5 million of annualized revenue potential.

The important word there is "potential". We would not add the full partnership figure to Klar's current revenue until there is evidence that it has actually converted into recurring sales. Still, the product has moved far beyond its first launch.

Klar's team has leaned heavily on universities, student communities, ambassadors and physical campus activations. Plinq followed another route. Founder Sabrine Matos built the Brazilian app to make public criminal information easier to access, and Lovable reported more than 10,000 users within three months alongside R$2.2 million in annualized revenue.

Neither product became useful because it was vibecoded. Klar gives students a clearer learning workflow. Plinq makes difficult public information easier to retrieve. Vibe coding mainly allowed small teams with very limited engineering resources to get those products into customers' hands quickly.

What are people actually paying these vibecoded apps for?

People currently pay vibecoded apps for very ordinary outcomes: making money, saving time, creating content, learning faster, running a business or changing a habit.

Look across the stronger examples and the variety is striking. Payout helps consumers recover money. Stoppr sells help with a personal habit. Stanley helps creators produce content. Klar helps students learn. QuickTables handles restaurant workflows. ShiftNex handles healthcare staffing. Lumoo produces fashion content. Plinq makes public safety information easier to access.

Pretty mundane, actually, and that's encouraging. It is a healthier picture than the early stereotype of vibe coding as thousands of interchangeable ChatGPT wrappers.

Customers rarely care how the application was implemented. A restaurant owner does not pay QuickTables because Lovable generated the React code. A Payout subscriber does not care whether Claude Code produced the backend. The build method matters enormously to the founder because it changes cost and speed. The customer still judges the result.

As software becomes easier to create, a weak problem becomes easier to turn into a polished weak product. The apps making money tend to start with something people already care about.

Is distribution now harder than building the app?

Yes. For vibecoded startups today, reaching customers increasingly looks harder than producing the first working version.

Payout built an influencer pipeline, learned which creator partnerships converted and then pushed successful creative into paid ads. Stanley had access to Stan's existing creator audience and also used public building, user interviews and targeted outreach. Klar went directly into student communities and campuses. Stoppr borrowed a consumer growth model from an app that had already proved demand in a neighboring category.

The build times are almost absurdly short next to the distribution work. Payout's first version appeared in days. Stanley was built in two weeks. QuickTables took roughly six weeks. Yet every one of those businesses still needed a separate answer to "where do the customers come from?"

That imbalance is becoming stronger as vibe-coding tools improve. Lovable currently says roughly 1.2 million new projects are created on its platform every week. Producing another application is clearly no longer scarce.

In that environment, distribution is the bigger edge. An audience, paid-acquisition skill, narrow-industry sales access or a real customer community gives a founder far more leverage than simply being able to build quickly.

How often do vibecoded projects actually fail?

A lot. The freshest public denominator we found comes from Pieter Levels himself, and his own project history is a useful antidote to the viral success screenshots.

Levels recently updated his public list of everything he has built. He currently labels nine projects as long-term successes, eleven as "Okay", nineteen as failures, six as too new to judge and dozens more as projects that were never intended to make money. He explicitly writes that most things he made never succeeded or made money.

The classification of fly.pieter.com is especially revealing. The flight simulator became one of the defining vibe-coding stories after reaching a $1 million annualized revenue run rate within 17 days. Levels now puts fly in the "Okay" group, which he defines as projects that took off and made money but were not sustainable.

That does not erase the launch. It gives us information we did not have when the screenshot went viral.

One of the most famous vibe-coded launches eventually failed its own founder's test for long-term success. That is the part launch screenshots hide. Levels still has nine projects in the success bucket, and his history shows why cheap experimentation is so valuable: founders can be wrong a lot when each experiment is cheap enough.

Does vibe-coded software fall apart once real customers use it?

Vibecoded apps can run real businesses, but recent security data shows that shipping AI-generated software without technical checks can become dangerous very quickly.

A recent Reeve study scanned 30,998 live apps built with platforms including Lovable, Bolt, v0, Replit and Base44. Among the 3,680 Supabase-backed applications whose databases the researchers could actually test from outside, 57% allowed an unauthenticated visitor to read at least one table.

Across the full sample, 1,332 apps, roughly one in 23, exposed some form of secret key in public code. Reeve also found 394 applications with readable tables whose names suggested personal or transactional data such as users, profiles or orders.

Those numbers need the right denominator. The 57% figure applies only to Supabase databases the researchers could reach, and Reeve performed automated external checks rather than full security audits. The same study found that 76% of the entire 30,998-app sample earned an A grade under its scoring system.

Separate academic work published this year has nevertheless found similar recurring problems in vibe-coded software, including placeholder logic, poorly filtered input, secret exposure and agents losing context as an application becomes more complex.

So scale changes the job. Getting the first paying users can happen with heavily AI-generated code. Once an app stores sensitive data, processes meaningful payments or serves thousands of customers, someone has to understand permissions, testing, backups, monitoring and security. AI can help with that work too, but revenue does not make engineering risk disappear.

Which vibecoded apps make money today?

Yes. Vibecoded apps are making meaningful money today, and Stanley, Payout and Stoppr have the freshest strong revenue evidence among the products we found.

Stanley currently gives us the largest recent figure, with roughly $3 million ARR reported across its creator products. Payout has unusually good verification because RevenueCat has observed the business from its early transaction data through its more recent growth to tens of thousands of dollars in monthly recurring revenue. Stoppr gives us a useful smaller-scale example: a first-time software founder reaching five-figure monthly revenue with a straightforward consumer subscription app.

Lumoo, ShiftNex, Klar and Plinq broaden the picture. Their public revenue numbers are less fresh and mostly come through founder or Lovable disclosures, so we would assign them lower confidence as statements of exactly what those businesses make now. They still show that prompt-built products have reached serious revenue in fashion, healthcare, education and consumer safety.

QuickTables adds an acquisition to the evidence. And, as we saw previously, fly.pieter.com now belongs in a different category: a spectacular vibe-coded launch that made real money but is classified by its own founder as commercially unsustainable.

At this point, the commercial question is settled: vibe coding can produce real software businesses. It has dramatically lowered the amount of engineering required to test an idea, and that change is visible in products reaching revenue within weeks rather than spending months getting to a first usable version.

The bottleneck has moved. Building another app is easy enough that Lovable users alone now create more than a million projects a week. Finding a problem worth paying for, reaching the right customers and keeping them interested remains much harder.

That is also why the best current vibecoding businesses look surprisingly normal. They sell restaurant software, healthcare staffing, content creation, learning tools, consumer subscriptions and access to useful information. Their founders found a cheaper way to build the software. They still had to build the business.