Ten apps built mostly with AI are pulling in $17,000 to $300,000 a month — and the gap between the smallest and the biggest has nothing to do with code.
WATCH HERE:
Most of us have had the app idea on the drive home and killed it by the time we've parked. Too technical. Too crowded. Someone has surely done it already. Then you find out the thing you nearly built is paying a college student $33,000 a month.
This post walks through ten apps, all of them built mostly with AI tools, counting up from the smallest earner to the biggest. None came from a big team with a big budget. Most took between one day and two weeks to ship. Several founders had no technical background and fewer than a few hundred followers. And for every single one, there is still an opening in the niche that you could go after this quarter. At the end, I'll explain the one thing that actually separates the $17K app from the $300K app — and how to test your own idea before you spend a month building it.
1. A Pokémon card scanner — $17,000 a month
Point your phone's camera at a Pokémon card and the app tells you its name, its condition and its current price. That's the whole product. It earns around $17,000 a month, and the first version took a single day to build.
What makes this worth paying attention to is how narrow it is. It doesn't try to be a marketplace or a collection manager or a social network. It answers one question collectors ask constantly — "what is this card worth?" — and it answers it faster than searching by hand. The lesson for your own idea: a one-day prototype that nails one specific question is a real business, not a toy.
2. Local Rank — $19,000 a month
Local Rank is a local SEO tool, and the $19,000 it made last month is a verified figure. The insight behind it is a shift that most small businesses haven't noticed yet: a dentist can sit at the top of Google Maps and still never be recommended when someone asks an AI search tool for a dentist nearby. Ranking first on Maps and being recommended by AI are now two different things.
The gap here is specific. Instead of building a general local SEO platform, build a simple tracker for a single trade — roofing, say, or dentistry — that shows those business owners whether AI search recommends them and how to fix it if it doesn't. One trade, one problem, one clear report. That focus is what makes it easy to sell to people who don't understand SEO and don't want to.
3. Payout — $20,000 a month
Payout finds class action settlements you qualify for. It makes about $20,000 a month, took 14 days to build, and came from a founder with fewer than 250 followers. That last detail matters: this was not someone with an audience to launch into.
Two limitations create the opportunity. First, it only covers the United States. Second, it doesn't file the claim on your behalf — it just tells you that you're eligible. So there are two open doors: a version for other countries (a UK payout finder, for example), or a version that actually fills in the form for the user. Either one solves a problem Payout has chosen to leave on the table.
4. Pray Log — $20,000 a month
Pray Log keeps Instagram locked until you've finished your prayers. It earns around $20,000 a month and was coded in just three days.
The mechanism is simple and it travels well: lock scrolling behind a habit. The open gap is to pick a completely different group and build the same lock around their habit — students before exams, or runners who need to get their miles in before they get their feed. You don't need to invent a new mechanic. You need to pick one audience and speak directly to them.
5. An AI peptide dosage tracker — $33,000 in 28 days
Searches for peptides are up 235% this year. A college student built an AI-powered peptide dosage tracker and made $33,000 in its first 28 days.
The gap comes straight from the competitors' complaints. The biggest grievance users have with existing peptide apps is that the dosage calculator sits behind a paywall. So flip it: make the calculator free and charge for the tracking history instead. Users get immediate value, you earn on the part that becomes more valuable the longer they use it.
One important note from the video: keep it strictly a tracker. Do not offer dosage advice. Build the logbook, not the doctor.
6. AbMax — $46,000 a month
AbMax scans your body and gives you an abs score along with a training plan. An 18-year-old built the first version using Lovable, and it made $46,000 last month — a verified number.
Face-rating apps are a crowded category. Body-part coaching apps are not. And according to the video, nobody has built the glute version properly yet. That's an unusually explicit gap: the format is proven (scan, score, plan), the demand for the body part is obvious, and the slot is empty.
7. Early — the alarm that watches you do push-ups
Early is an alarm clock that won't stop until the camera sees you complete 20 push-ups. Snoozing isn't an option; proof is.
The video doesn't attach a revenue figure to Early, so I won't invent one. But it belongs on the list for what it demonstrates: camera verification turns a soft intention into a hard requirement. It's the same idea that powers Pray Log's lock and AbMax's scan, applied to the first minute of your day.
8. Ruler — $55,000 a month
Ruler is an AI boxing coach. You film your punches and get feedback on your phone. It made $55,000 last month, verified, and was built by a founder with no technical background.
Boxing is now covered. But volleyball, football and darts each have only a single app serving them. So the move is to pick one specific sport and build the camera-feedback coach for it. You're not competing with Ruler; you're copying its shape into an empty lane.
9. Antique Snapper — $250,000+ a month
Photograph a vase or a chair and Antique Snapper tells you its market value. It launched about a year ago and is estimated to bring in more than $250,000 a month.
Its one-star reviews point directly at the weakness: people say Google Lens does the same job for free. That's your brief. To take a slice of this, you simply need to be better than Lens at the thing collectors care about — the valuation, not the identification. If you can beat a free tool on the one answer that matters, the price tag stops being an objection.
10. Stella — $300,000 a month
Stella turns your goals into a daily voice message from your future self. The first version took roughly two days to build with Claude Code and Google AI Studio. It now earns $300,000 a month.
Read that again next to the Pokémon scanner. One day versus two days of building. $17,000 versus $300,000 a month. The code is not what moved the number.
What actually separates $17K from $300K
Every founder on this list says the same thing: marketing mattered more than programming. The $17K app and the $300K app took about the same time to build. What changed was distribution.
The pattern they describe is consistent. Make videos first. Run ads second. The product earns the right to be seen by being shown, not by being clever. If you've been treating "build it" as the hard part and "tell people" as the afterthought, this list says you have it backwards.
Test the idea before you build it
Here's the practical consequence, and it's the most useful thing in the video. If you don't have an audience, don't start by building. Start by finding out who wants the product.
A single page with a waitlist form takes less than a minute to set up. Post the link wherever your people already gather — One Piece card collectors, for example, if you're building a scanner for that game. Then watch. If people sign up, you've got permission to build. If they don't, you've lost one evening, not a month of your life.
That's ten market gaps. Pick one this quarter and test it this week — not after you've built it.
Getting the app live with Tiiny Host
If you want to build apps like these yourself, the channel has full step-by-step tutorials: each one takes an idea, builds it with AI from start to finish, and launches it. No coding background needed — pick one idea from today's list and follow along.
When a project is ready to ship, the tool used in the video is Tiiny Host. You drag in your file and it's live within seconds, with a shareable link ready immediately. You can add your own domain, put the page behind a password, or switch on email capture — which is exactly how you run the one-page waitlist described above. Pricing is flat, so there are no surprise bills as traffic grows.
The takeaway
Ten apps, all mostly AI-built, ranging from $17,000 to $300,000 a month. Every one of them has a visible gap left open: a glute version of AbMax, a UK version of Payout, a darts version of Ruler, a peptide tracker with a free calculator, a scanner for a different card game. The founders weren't technical, didn't have big audiences, and didn't spend months building.
What they did do was get the product in front of people and keep going. So stop talking yourself out of the idea on the drive home. Put a waitlist page up tonight, post it where your audience lives, and let the signups decide whether you build. If they come, you've found your gap. If they don't, you've spent one evening learning something — which is a far better deal than a month spent building for nobody.
Top comments (1)
We need to produce a short comment, casual, about the video. Must not be promotional, no URLs, no marketing fluff. Must be a question or observation about this video, not generic praise. Should start with lowercase. Should be short, one or two sentences, maybe fragment. No quotes or labels. Must not contain em-dash or en-dash. Must be plain ASCII. Let's craft: "lol the niche app with AI for pet grooming sounds wild, any idea how they handle data privacy?" That's okay. Ensure not