ProvenStartups
IdeasPricingMethod
Get access
ProvenStartups

Startup ideas with revenue receipts, reverse-engineered from founder interviews.

contact@provenstartups.com
Product
  • All ideas
  • Pricing
  • Method
  • Blog
Company
  • About
  • FAQ
  • Contact
Legal
  • Privacy
  • Terms
  • Refunds
Official accounts
  • LinkedIn
  • GitHub
  • Reddit
  • Pinterest
© 2026 ProvenStartupsNo fabricated numbers. Ever.
Home/Blog/Built With AI

Best Ways to Make Money With AI, Ranked by Documented Revenue

We rank by the strength of the evidence behind each revenue figure, not by how good the idea sounds. Of the 106 products built with AI coding tools in our…

ProvenStartups·Published 2026-08-01

We rank by the strength of the evidence behind each revenue figure, not by how good the idea sounds. Of the 106 products built with AI coding tools in our index, 32 disclose a clean monthly number, the median is $25K/mo, and the range runs $37/mo to $2M/mo.

State the sort order before the list, or it is taste with numbers glued to it. Here is ours.

Table of contents

  • ·What are the best ways to make money with AI?
  • ·The ranking, and the evidence behind each entry
  • ·How much is each of these numbers actually worth?
  • ·Why the usual rankings are wrong
  • ·What sits in the middle of this ranking
  • ·How to use this without fooling yourself
  • ·Frequently asked questions

What are the best ways to make money with AI?

The ones where a figure survived contact with a document. Of 106 indexed products built with AI coding tools, only 32 disclose a clean monthly figure — median $25K/mo, floor $37/mo, ceiling $2M/mo. A verified $440K/mo outranks a relayed $1.3M/mo, every time.

This group is better evidenced than our index overall, worth knowing before someone tells you AI revenue claims are uniformly fictional. They are not uniformly anything.

One limit we apply to ourselves: the index over-represents businesses that succeeded and then chose to disclose. Nobody publishes a teardown of the wrapper that made $37 a month, though $37/mo is this dataset's floor. The median judges someone else's claim; it does not forecast yours.

A focused individual types on a laptop running AI software indoors.
Photo by Matheus Bertelli on Pexels

The ranking, and the evidence behind each entry

Sorted by evidence class first, then by size.

BusinessDisclosed figureEvidence classTeam
The Viral App Monetization Machine (100 Apps Dissected)Cal AI & Lerna $2M/mo each · LazyFit, CoinSnap & Impulse $700K/mo eachThird-party verifiedSmall teams across the board
Niche Identifier App Portfolio (CoinSnap as the Model)$500K/mo · top identifier app (Sensor Tower)Third-party verifiedCompany, portfolio operator
Gravl$440K/mo · 70K+ subscribersThird-party verifiedStarted with Julian and two others, now 13–14 people
Cluely$500K/mo (~$6M ARR)Founder-reportedRoy Lee + co-founder Neil, ~13–14 people plus 60 contracted creators
Grower$180K/mo MRR · ~10% month over month · 50+ employeesFounder-reported50+ employees (Barış Korkmaz, Istanbul, Turkey)
AI Clone of Floe-Style Agency SaaS$1.3M/mo for the original benchmark; the cloner himself, $0Creator-relayed1 person, a build-demo creator
Upward$400K/mo, from revenue-tracking toolsCreator-relayedNot disclosed in the video
TinyURL Clone Link Shortener (Replit + AdSense)Benchmark TinyURL: $200K/mo · 40M monthly visitsCreator-relayedSolo (Mikey, non-coder)
Receipt-Scanning AppsSimplyWise estimated at $60,000/month; a third app at $80,000/monthCreator-relayedNot disclosed
AI-Built Single-Page Read-Later App$60K/mo for the benchmark appCreator-relayed1 person, a build-demo creator

Grower is the entry to study. Alongside the $180K/mo MRR the founder discloses 200+ companies applying to sign up daily, 6–7% of customers in London, and monthly revenue at 3x total investment ever raised. Nothing else in the table discloses at that density.

How much is each of these numbers actually worth?

Within this group the evidence splits 20 third-party verified, 49 founder-reported, 20 creator-relayed, 17 unproven — 65% verified or founder-reported. Better than the index as a whole; still one number in three arrives secondhand or with nothing behind it.

Evidence classCount in this groupWhat it means
Third-party verified20Checked against a document, dashboard or independent dataset. Gravl's $440K/mo sits here.
Founder-reported49The operator states a figure about their own business. Grower's $180K/mo MRR sits here.
Creator-relayed20A third party reports a number about a business they do not run. Highest drift.
Unproven17A claim with no support offered at all.

This is the argument the whole site rests on: a founder-reported figure and a verified figure are not the same claim, and quoting them with equal confidence is the standard failure of AI-money content. The internet keeps the number and drops the class; we keep both — how we grade the evidence.

Compare the whole index: 406 businesses, 106 disclosing, median $27K/mo, split 57 verified / 184 founder-reported / 121 creator-relayed / 44 unproven. App-store revenue is partly observable from outside, which is why this group scores better.

A man working on a laptop in a cozy, modern office space with a focus on technology.
Photo by Matheus Bertelli on Pexels

Why the usual rankings are wrong

Most "best ways to make money with AI" lists sort by search volume or by whatever excited the writer, floating the loudest unverified claim to the top. Sorting by evidence drops the $1.3M/mo headline to sixth and puts three verified entries first.

Three distortions our sort removes:

  • ·Benchmark laundering. A build demo quotes an incumbent's revenue while shipping a clone, and readers credit the clone. Two entries above do this; one states $0.
  • ·Tool estimates dressed as disclosures. Sensor Tower output and a founder's own P&L are different objects. Upward's $400K/mo comes from revenue-tracking tools, and every receipt-scanner figure is an on-screen App Store estimate — a screenshot of a guess.
  • ·No failure column. We keep 7 documented failures here and 38 across the index. A ranking that cannot name a loss is a marketing artefact.

Two of those failures teach more than the wins. AI Co-founder Club reached $8,374 MRR from 270 paying customers 13 days after launch — and the product is not AI, just a monthly fee for putting a scary tool in a safe cage. Quick Shorts made $500+ on launch day off one 2.3M-view Reel, plus nearly 10K emails and a 1,000-person Discord, built in Lovable in a week once demand was proven. The build was the cheap part.

For base rates, the U.S. Census Bureau's Business Trends and Outlook Survey measures adoption directly and the Stanford HAI AI Index Report covers capability and investment. Both beat a creator's assertion about market size.

What sits in the middle of this ranking

The median is $25K/mo, and that is where a reader copying any of this lands. 71 of the 106 businesses here are one-person operations — unlike the top of the table, where Gravl runs 13–14 people and Grower runs 50+.

Middle entries look alike: one product, one named buyer, a price tested against churn, a channel the founder can name in a sentence. NoFap / Men's Self-Improvement App reports $6K/mo in month one with about 1,100 paying users, on positioning alone — sell the outcome, not the prevention of a problem.

And the floor is $37/mo — a real indexed attempt, same tools, same year.

A person using a laptop to interact with AI technology indoors during the day.
Photo by Matheus Bertelli on Pexels

How to use this without fooling yourself

Four checkable moves this week.

  1. 1.Grade the claim that brought you here. Only 20 of 106 here are third-party verified, so the base rate says yours is founder-reported or weaker.
  2. 2.Plan against $25K/mo, stress-test against $37/mo. Then read the 7 documented failures here as carefully as the wins, from all indexed ideas.
  3. 3.Prove demand before you build. Quick Shorts proved it with one Reel before writing the product. More demand-first patterns in Way To Make Money Online and Ways To Make Money Online For Women.
  4. 4.Write the one-pager, then pick three comparables. Structure from the SBA's guide to planning a business; comparables in the Built With AI hub. Read team size and evidence class before revenue.

Frequently asked questions

How to make $1000 a day with AI?

Our data contains no verified $1,000/day playbook, and we will not invent one. The nearest reference is Quick Shorts at $500+ on a single launch day, after a Reel with 2.3M views produced nearly 10K emails. The floor of this group is $37/mo, and 7 of its 106 entries are documented failures.

How to make $10,000 a month on the internet?

$10K/mo sits below the $25K/mo median of the 32 disclosing AI-built products, so a majority of the disclosing set clears it. A realistic small case is the men's self-improvement app at $6K/mo in month one with about 1,100 paying users, founder-reported. Remember the index skews toward businesses that succeeded and chose to disclose.

Can I use AI to make me rich?

The top of this group is $2M/mo each for Cal AI and Lerna, third-party verified. But 71 of the 106 businesses here are one-person operations sitting nearer the $25K/mo median, and the $1.3M/mo headline in our table belongs to a benchmark whose cloner reported $0. The tail is real and rare.

How reliable are these best ways to make money with ai figures?

Stronger than most categories we index, and still uneven: 20 third-party verified, 49 founder-reported, 20 creator-relayed, 17 unproven — 65% verified or founder-reported. Discount hardest on tool-derived estimates, like the receipt-scanner apps at $60,000/month and $80,000/month, all shown from an unnamed source tool.

How many of the 106 indexed businesses actually disclose a monthly number?

32 of 106. The other 74 are indexed on evidence that never included a clean monthly figure. Across the whole index the ratio is comparable: 106 of 406 businesses disclose a clean monthly number, median $27K/mo, spanning 22 countries, with 38 documented failures kept on record.

← More in Built With AIBrowse proven ideas