Turns any song into an AI-voiced marble race video in the browser, which can be scheduled straight to TikTok, Shorts and Reels.
The TikTok/Reels format of "marble race + a song + each marble singing a line in an AI voice" took off in Latin America (especially Mexico and Argentina), and RaceCut registered its domain two months ago and within weeks built a browser tool where you "drop in a song -> it automatically races, syncs to the beat, adds AI voices -> outputs a vertical MP4," grabbing the tooling window for this format. The Spanish-language creator account dancoversia has kept posting "made with @racecut.app, link in bio" videos since 2026-08 and co-posts with the official account; people who see the videos search the brand name to arrive (brand keywords are 74% in our data) and convert on the low-barrier $1/$5 first-month subscription.One-line growth story, compiled by us from public sources; speculation is marked separately below.
The main engine is going viral: marble-race singing videos took off on Spanish-language TikTok, and people who saw them searched for racecut and visited the site. Organic search is second and is not paid; the top keywords are almost all branded.
Spanish-speaking creators promoting the tool act as the turbo boost: dancoversia keeps posting "made with @racecut.app, link in bio" and how-to tutorials, and the official Instagram posts are co-branded with him.
By our estimate, about two thirds of visits come from people who saw a video or tutorial; visit counts are estimates, and about 23 of every 100 visits click through to the checkout page.
Model: Lenny's racing growth framework. Engine = going viral (a video format that is already popular: see a marble race video, search racecut, make one yourself and post it); turbo boost = Spanish-speaking creator dancoversia promoting it and posting tutorials; lubricant = a low-barrier subscription at $1/$5 for the first month, plus creators who need a new video every day coming back to make the next one.
Real data: Month-over-month +2811%, the largest channel is direct at 35.65%, top three channel ranking (direct, organic search, referrals), organic search is 100% of search, 6 search keywords in total, top keywords (racecut, racecutapp, racecut app, marble race ia, racecut.spp), countries (Mexico 67.36%, Argentina 32.64%), bounce rate 11.45%, 3.29 pages per visit, average stay 1 minute 31 seconds, global rank #821240, all from the Similarweb public page for last month; the public page does not give total visits or the shares of the second and third channels, and referral and social sources are empty. The 8,596 visits (about 8,600) to the Stripe checkout page come from quietmoney checkout data.
Supporting evidence: racecut.app was registered in 2026 and launched within a few weeks; the marketing page says "5,000+ creators" and "As seen all over TikTok"; dancoversia posts "made with @racecut.app" videos and Spanish tutorials on Instagram and TikTok, there are also racecut.app videos on YouTube Shorts; the official Instagram followers grew from about 3,000 to 6,472, with posts co-branded with Dancoversia. The founder was not found.
Inference: The 37,600 total visits are estimated from the Similarweb global rank #821240 (fitted on 437 sites with real traffic, half of sites within 17% error), so both channel circles carry ≈; the channel shares and rank are real. How each route splits is inference: direct 13,400 = creators coming back 8,400 + opening directly after seeing a video 3,500 + clicking the creator's bio link 1,500; other channels 24,200 = searching racecut after seeing a video 15,000 + creators searching the brand name to come back 4,700 + creator tutorials and links 4,500. The 2% visit rate after seeing a video and the 5% click rate after watching a creator tutorial are guesses; whether videos posted by users carry a watermark was not verified. In the quietmoney data, more than 100 sites show exactly 8,596 visits last month, which looks like the data source estimating in bands, so real checkout page visits may differ.
Checkout: checkout.stripe.com
“'Connected TikTok, Shorts and Reels accounts. Schedule from RaceCut, no downloads.'; Unlimited plan includes 'no watermark' (implying watermark on other output)”Quote from the site
New-user source: Viral (Built-in sharing). Search 48.3% (brand keywords like racecut 74.3%, non-brand such as marble race ia), direct 35.6%, referral 9.7%. New customers mainly see a watermarked marble race video on TikTok and then search the brand to arrive, while returning users come straight back to make the next one.
Top channel measured by Similarweb, others estimated by rank
| Attributable new users, last month | Share |
|---|---|
| SEO | 10% |
| 74% | |
| Other | 16% |
Direct + brand search 74%. 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 |
|---|---|
| Mexico | 67.36% |
| Argentina | 32.64% |
Mainly English, also Spanish
6 keywords · brand ~74% · organic 100% / paid 0%
| Similarweb top 5 keywords |
|---|
| racecut |
| racecutapp |
| racecut app |
| marble race ia |
| racecut.spp |
Non-brand queries: marble race ia
About 8,596 last month (Similarweb estimate), #364 of 606; the month before was 207. Opening checkout is not paying, so this is not revenue.
New-user source: Viral (Built-in sharing). The site has invite or share mechanics, and direct plus brand search is the largest source Search 48.3% (brand keywords like racecut 74.3%, non-brand such as marble race ia), direct 35.6%, referral 9.7%. New customers mainly see a watermarked marble race video on TikTok and then search the brand to arrive, while returning users come straight back to make the next one.
$1 for the first month, then $24.99/month; the page also has a $6.99/month tier; non-unlimited plans carry a watermark
Sites in the same category: Mage, Emochi, Ssemble, Polarisugc. Each has its own page here with revenue or checkout visits, traffic sources and how it started.
People do pay for it: about 8,596 visits went from the site to its checkout page last month (Similarweb estimate). We have no affiliation with it and did not test the product.
Dossier confidence: medium. 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.