Pick a city on a 3D globe, set the yield and burst height, and simulate a nuclear detonation's blast radii, thermal radiation, simplified fallout and estimated casualties (using WorldPop population data).
A solo mini-site registered on 2026-08-25 and launched about a week later: search any city on a 3D globe, pick a bomb (Little Boy, Tsar Bomba, W87, etc.) and a burst height, and watch the fireball, blast rings and WorldPop-based casualty estimates. It looks far better than the venerable NUKEMAP, and screenshots and recordings of "dropping a nuke on my own city" are naturally shareable. In early last month it got reposted across X, Threads and Facebook; WorldStar reshared a screen recording under the headline "new nuke simulator supposedly built from a single AI prompt has people wildin" (about 69K views), and from there it spread to the Korean community aagag.com, Spanish-language tech blog microsiervos, Brazil and elsewhere. Every simulation result is a share link carrying coordinates and weapon parameters (?lat=…&weapon=…), so each click is a new visit. The author then added SEO landing pages for cities, weapons and nuclear test sites to catch long-tail search, and monetizes with weekly/monthly/quarterly Pro passes sold via Polar (unlocking photorealistic 3D cities).One-line growth story, compiled by us from public sources; speculation is marked separately below.
The main engine is viral spread: every simulation is a share link carrying city and weapon parameters, and screen recordings of "nuking your own city" are naturally shareable. Direct visits are 38%, and the search terms are all brand terms.
The turbo boost was a burst of reposting: WorldStar reposted a screen recording that got about 69,000 views, which then spread to the Korean community aagag.com, Threads and the Spanish blog microsiervos.
By estimate, about half of visits come from people who saw someone else's nuke recording or post; visits are estimated; about 0.5 of every 100 visits click through to the checkout page.
Model: Lenny's racecar growth framework. Engine = viral spread (the output carries the brand: every simulation is a shareable parameter link and recording, and when someone opens it that is a new visit); turbo boost = a burst of reposting (WorldStar, Korean communities, Threads, Spanish blogs); lubricant = the leaderboard and city and weapon landing pages, which turn one-time onlookers into repeat visits and search.
Real data: Biggest channel direct at 38.3%, top-three channel ranking (direct, organic search, organic social), organic search 100% of search, 112 search terms, top search terms (nukesimulation, nukesimulation com, nuke simulation com, nuke simulation, nukesimulation.com), top referrals (aagag.com 7.77%, threads.com 6.57%, microsiervos.com 4.84%), social sources (X, Facebook, Reddit), countries (US 25.75%, South Korea 13.16%, Brazil 11.16%, India 8.15%, Turkey 5.13%), bounce rate 37.62%, 34.77 pages per visit, average stay 2 minutes 43 seconds and global rank #22368 come from the public Similarweb page for last month; the public page does not give the shares of the second and third channels. 10,481 visits (about 10,500) to the Polar checkout page come from quietmoney checkout-page data.
Supporting evidence: The domain was registered in 2026 and launched about a week after registration; it is a small one-person site with no name on the official site. On launch day Wayback already saved share links with city parameters; WorldStar reposted a recording that got about 69,000 views; about 10 days after launch it added landing pages for cities, weapons and nuclear test sites, and later added a leaderboard and account endpoints; it sells one-week, one-month and three-month Pro passes through Polar.
Inference: The total of 1,953,400 visits is estimated from Similarweb global rank #22368 (fitted on 437 sites with real visits, where half of the sites are within 17% error), so the two channel circles are estimates too; only direct at 38.3% and the channel ranking are real figures. How each path is split is an inference: direct 748,100 = links from friends 350,000 + coming back to simulate again 248,100 + seeing recording posts 150,000; other channels 1,205,300 = searching the name or clicking a link after seeing a recording post 800,000 + friend links 150,000 + coming back to simulate again 150,000 + searching long-tail city and weapon terms 105,300. The ratios of 5% visiting after seeing a recording post, 40% opening after receiving a friend's link and 10% clicking in from long-tail search are guesses. The old channel numbers in the profile (search 46.3%, direct 38.3%) use an earlier basis, so only the public page's ranking is used here.
Checkout: polar.sh
New-user source: Social media (Organic social). Search 46.3% (brand keywords 84.6%, nukesimulation etc.), direct 38.3%, social 9.3% (X, Facebook, Reddit). Referrals from the Korean community aagag.com 7.77%, threads.com 6.57%, Spanish-language tech blog microsiervos.com 4.84%; spreads by being reposted in communities.
Top channel measured by Similarweb, others estimated by rank
| Attributable new users, last month | Share |
|---|---|
| 9% | |
| Other | 91% |
Direct + brand search 85%. Direct and brand search count as viral only when the site shows sharing mechanics; the largest share wins, under 5% means Other.
| Country | Share of visits |
|---|---|
| United States | 25.75% |
| Korea, Republic of | 13.16% |
| Brazil | 11.16% |
| India | 8.15% |
| Turkey | 5.13% |
English
X (Twitter), Facebook, Reddit
112 keywords · brand ~85% · organic 100% / paid 0%
| Similarweb top 5 keywords |
|---|
| nukesimulation |
| nukesimulation com |
| nuke simulation com |
| nuke simulation |
| nukesimulation.com |
23 referring sites
| Site | Share of referrals |
|---|---|
| aagag.com | 7.77% |
| threads.com | 6.57% |
| microsiervos.com | 4.84% |
| Category | Share of referrals |
|---|---|
| Computers Electronics and Technology - Other | 27.41% |
| Video Games Consoles and Accessories | 17.24% |
| News & Media Publishers | 10.23% |
| Ecommerce and Shopping - Other | 9.61% |
| Social Media Networks | 6.41% |
About 10,481 last month (Similarweb estimate), #321 of 606; the month before was 12. Opening checkout is not paying, so this is not revenue.
New-user source: Social media (Organic social). New users mostly come from YouTube, X, Reddit, TikTok and the like Search 46.3% (brand keywords 84.6%, nukesimulation etc.), direct 38.3%, social 9.3% (X, Facebook, Reddit). Referrals from the Korean community aagag.com 7.77%, threads.com 6.57%, Spanish-language tech blog microsiervos.com 4.84%; spreads by being reposted in communities.
Core features free; Pro is a time-limited pass by week/month/three months (paid via Polar), amount not found
Dossier confidence: high. Parts 1 and 2 come from reading the site and Similarweb data; part 3 from public reporting, founder posts, Wayback snapshots, WHOIS and similar, with speculation marked.
The number: checkout visits is Similarweb's estimate of visits from the site to a checkout payment page; opening checkout is not paying, so this is not revenue. Small sites land on fixed buckets.
Data through last month. Page updated 2026-10-31.