MONEY LIST · 10 receipts
AI apps built in a weekend that make money
This list starts with build time, not a promise of speed. Each matching receipt says that the product took hours, a weekend, or a few days to reach the stated outcome. That is a useful filter because it shows what makers shipped quickly, but it does not turn a short build into a repeatable formula. Some records are revenue, one is a sale, some are savings, and one is a prize. The kind stays visible in the receipt so those outcomes are not mixed together.
Fly is the largest recurring-revenue example in this set: Pieter Levels reported $12,000 MRR after an initial prototype that took about three hours. Frank Michael Smith reported $40,000 in monthly revenue for GeoSports, built in eight hours over a weekend. TrustMRR reports $17,238 MRR after about six hours for its first version. LaunchFast reports $20,000 MRR after a 48-hour MVP, while Giovanni Brees reports $45,000 MRR for MeetOscar after building its first version over a single weekend.
The other outcomes show why the number needs its label. Josh Pigford's NameSnag took about 12 hours to build and has a reported $15,000 acquisition price. SPED reports a $50,000 hackathon grand prize after 24 hours. Jon Cheney's GenAIPI reports $105,000 of agency cost avoided after a weekend. Martin Ebongue reports $96,000 of annual creative-studio cost avoided after one afternoon of setup and a few days of learning. Max Votek reports a $2,000 monthly SaaS bill replaced by a personal CRM built in about 20 hours.
There is a common pattern in the descriptions: the first version has a narrow job. Fly is a simulator, GeoSports is a single game loop, LaunchFast focuses on product sourcing, MeetOscar drafts email replies, and NameSnag searches for available domains. The receipts do not say that these products started as full platforms. They describe a small first use case, then a later revenue, sale, saving, or award claim. That is the practical lesson to test: make the first job clear enough to ship, then see if people pay or if a real cost disappears.
The evidence is mixed. TrustMRR is maker-documented, while GeoSports, LaunchFast, MeetOscar, and Max Votek are maker claims. The figures may describe MRR, monthly revenue, a sale price, a prize, or avoided spend. Read the linked receipt before comparing them. This is a shortlist of fast builds with reported outcomes, not proof that any app can make money in a weekend.
Key takeaways
- A fast first version can support revenue, savings, a sale, or a prize, but these are different outcomes.
- The clearest examples start with one narrow job, such as domain search, email drafting, or product sourcing.
- MRR, monthly revenue, acquisition price, prize money, and avoided cost are not interchangeable.
- Most figures are maker claims; read the evidence label and source before treating one as a benchmark.
- The receipts support testing a small first use case, not copying a weekend revenue promise.
Receipts in this list
| Product | Maker | Amount | What it is | Built with | Evidence |
|---|---|---|---|---|---|
| GenAIPI | Jon Cheney | $105,000agency quote avoided | GenAIPI is an online school that trains business teams to use AI tools. | Replit | Maker-documented |
| martinebongue.com content studio (five free Claude skills) | Martin Ebongue | $96,000annual creative-studio cost avoided | A solo creator replaced his five-person content team with five free Claude skills. | Claude, NotebookLM skill, Nano Banana, Remotion skill, Stitcher skill, Humanizer skill | Maker claim |
| SPED (Training-Free Interactive World Models) | Bryan Dong, Ray Kasichainula, Subha Vadlamannati and Gino Chiaranaipanich | $50,000grand prize, Inference-Time Compute Hackathon 2026 | SPED makes autoregressive video diffusion interactive, letting a user retarget a generating video mid-stream at real-time speed. | Gemini, FlashAttention 3 | Maker-documented |
| MeetOscar | Giovanni Brees | $45,000reported MRR | MeetOscar is an AI email assistant that drafts replies in the user's own voice. | Lovable, ChatGPT API | Maker claim |
| GeoSports | Frank Michael Smith | $40,000reported monthly revenue | GeoSports is a daily sports geography web game where players tap the map where a moment happened. | Claude | Maker claim |
| LaunchFast | Hassam | $20,000reported MRR | LaunchFast finds, validates and sources Amazon private-label products in minutes. | Cursor | Maker claim |
| TrustMRR | Marc Lou | $17,238Stripe-verified MRR | TrustMRR lists founders’ Stripe-checked revenue and grew into a marketplace for buying and selling small online businesses. | Cursor | Processor-verified |
| NameSnag | Josh Pigford | $15,000acquisition price | NameSnag scans expired .com domains to help users find available names for new projects. | Claude Code | Maker claim |
| Fly | Pieter Levels | $12,000reported MRR | Fly is a free multiplayer flight simulator monetized with an in-app plane and ads. | Cursor AI, GPT-4, Grok 3, Claude 3.7 Sonnet | Maker-documented |
| Personal CRM built with Claude Code | Max Votek | $2,000monthly SaaS bill replaced | A solopreneur replaced HubSpot, Monday.com, Notion and Zapier with a CRM he built in Claude Code. | Claude Code | Maker claim |
FAQ
What qualifies for this list?
A receipt qualifies when its stated tookTime mentions hours, a weekend, or a few days. The page uses the matching receipt IDs because the list filter cannot query tookTime directly.
Are all of these apps built in exactly two days?
No. The receipts use different wording, including about three hours, 48 hours, a weekend, 24 hours, and a few days. Read each tookTime field for the exact claim.
Does a short build time prove the product was easy?
No. The receipts describe a first version or a stated build period. They do not measure later product work, distribution, support, or maintenance.
Are the reported amounts verified revenue?
No. The list includes maker claims, maker-documented figures, a public-record prize, and a reported sale. The evidence column and linked receipt show the limit.