AI influencer generation studio: set up a virtual person with a fixed look, then generate selfies, outfit photos and short videos of her in bulk.
A US (Florida LLC) AI influencer studio whose domain was only registered three months ago, riding the "faceless account / AI influencer side hustle" wave: users set up a virtual girl with a fixed look and generate her mirror selfies, outfits and short videos in bulk, with credit packs from $3 and no forced subscription lowering the barrier to try. In 1–2 months ago, "AI influencers are PRINTING / side hustle" bloggers on Instagram named dirtybunny.ai in their Reels, bringing the first users, and after that brand-term search and returning users spending credits keep it going. About 3 months after launch it claims 7,000+ (elsewhere 42,000+) creators and 139,000 images.One-line growth story, compiled by us from public sources; speculation is marked separately below.
The main engine is side-hustle creators naming it and driving traffic (inference): side-hustle creators on Instagram who talk about making money with AI influencers, and early users, name it in their posts, and viewers then search the brand name; the biggest channel is organic search at 55%, and the top terms are all branded.
The lubricant is the credit packs starting at $3 that do not expire: the barrier is low, and returning users open the site directly and keep generating with their credits; visitors who came by asking an AI assistant ranked in the top three channels, and last month's visits rose 107% month over month.
By estimate, more than 40% of visits are returning users who keep generating with their credits; visits are estimated, and about 25 in 100 visits click through to the checkout page.
Model: Lenny's racecar growth framework. Engine = side-hustle creators naming it and driving traffic (inference: creators on Instagram who talk about making money with AI influencers and early users give the tool name in their posts, and viewers search the brand name); lubricant = credit packs starting at $3 that do not expire, which lower the barrier to trying it, with returning users coming back to keep using their credits.
Real data: The biggest channel, organic search at 54.94%, the top-three channel ranking (organic search, direct, AI assistants), organic search at 100% within search, 6 search terms in total, the top search terms (dirtybunny, dirty bunny, dirty bunnyai, dirtbunny ai, dirtyubunny ai), countries (US 100%), bounce rate 46.06%, 5.37 pages per visit, average time on site of 4 minutes 7 seconds, month-over-month +106.7% and global rank #624455 come from the Similarweb public page, last month; the public page does not give total visits, does not give the shares of the second and third channels, and does not list social sources. 12,894 visits (about 12,900) to the Stripe checkout page come from quietmoney checkout page data, and it was 0 in the earlier months.
Supporting evidence: The domain was registered in 2026, with an SSL certificate issued the same day; the operator is Dirtybunny LLC in Florida, and the founder is not public; at least two Instagram posts name dirtybunny.ai in the caption, one of them a "making money with AI influencers" Reel (344 likes) by side-hustle creator Hakeem Haddon; the homepage claims 7,000+ creators (42,000+ elsewhere) and 139,000 images, and the figures contradict each other.
Inference: The 50,700 total visits is estimated from Similarweb global rank #624455 (fitted on 437 sites with real visits, with half of the sites within a 17% error), so both channel circles carry an approximate sign, while the share of the biggest channel is real. How each route splits is an inference: search 27,900 = searching the brand name after seeing Instagram posts 19,000 + returning users searching the brand name 8,900; other channels 22,800 = returning users opening directly 13,000 + opening directly after seeing posts 6,000 + asking AI assistants 3,800. The rates of 2% visiting after seeing a post and 20% clicking in after an AI assistant recommendation are guesses; whether the side-hustle creators are paid or affiliate promoters was not found. The profile's direct 33.8% and AI assistants 6.8% use an old basis and are not used here; the profile marks these 12,900 visits as an estimate, and 12,894 is identical to several other sites in the same batch, so it may be an estimated lower bound, and it is used as is here.
Checkout: checkout.stripe.com
“See the full gallery →; Real outputs — every one generated in the studio, not a stock photo; looks like a real phone photo, not AI”Quote from the site
New-user source: GEO (AI assistants). Search 54.9%, direct 33.8%, AI assistants 6.8%; search is almost entirely brand terms (dirtybunny, dirty bunny, brand keywords 88.7%), and new users are classified as GEO at about 6.8% from AI assistant referrals. Returning users come back directly to use their credits.
Top channel measured by Similarweb, others estimated by rank
| Attributable new users, last month | Share |
|---|---|
| GEO | 7% |
| Other | 93% |
Direct + brand search 89%. 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 | 100% |
English
Gen AI is #3 channel, no % shown
6 keywords · brand ~89% · organic 100% / paid 0%
| Similarweb top 5 keywords |
|---|
| dirtybunny |
| dirty bunny |
| dirty bunnyai |
| dirtbunny ai |
| dirtyubunny ai |
About 12,894 last month (Similarweb estimate), #234 of 606; the month before was 0. Opening checkout is not paying, so this is not revenue.
New-user source: GEO (AI assistants). New users mostly come from ChatGPT and other AI assistants Search 54.9%, direct 33.8%, AI assistants 6.8%; search is almost entirely brand terms (dirtybunny, dirty bunny, brand keywords 88.7%), and new users are classified as GEO at about 6.8% from AI assistant referrals. Returning users come back directly to use their credits.
Credit packs bought one-time: $3/50, $5/100, $10/250, $20/500, $70/2,250 credits; no subscription needed
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.