
2 thousand installs, #11 top grossing in Italy: Dudufi is the team's 5nd reskin
What it is An Android app: pick a template, upload a photo, and get a stylized image or a 8–15 second AI video. Previews are free; generating requires buying credits or a membership.
Why it took off It uses ByteDance's BytePlus video model API off the shelf, charges from the very first generation, and feeds payment data back to Meta so it only finds people willing to pay. It sits in the Video Players & Editors category, where the bar to chart is low, so 2 thousand installs was enough to reach the top grossing chart.
BHow you could play it A small team that already runs Meta app ads, connects to a per-second video model API, picks one concrete niche (pets, weddings, babies) for its templates, uses a hard paywall plus payment event feedback, and tests with a small budget in cheaper secondary markets first.
Est. monthly revenue (Google Play) $1K–10KWe estimated this from its rank on country top-grossing charts and real install counts, calibrated against real revenue; typical error is 2–3x. Read it as a tier, not an exact number.
- Data (updated daily) · first launch 2026-10-23 · 11-19 installs 7,052
Android · Total installs worldwide Data 2026-11-10 → 2026-11-19 (10 days, 7 observations). Google's detail page only refreshes install counts every other day, so a flat day doesn't mean nobody installed.
Dudufi is an Android app built on one idea: upload a photo, apply a template, get an AI short video. Three weeks after launch it had 2,247 installs worldwide, no ratings, and was already #11 in the Video Players & Editors top grossing chart in Italy. Behind it is a playbook: the same code and the same backend, launched under 5 names across two months from 5 developer accounts. What's worth studying is the playbook, not this app.
A Utah household goods company launched an AI video app
On Google Play, Dudufi's developer is JNH QuickShop, with an address at a house number in Centerville, a small town north of Salt Lake City, and the contact email [email protected]. The .buzz domain is an online grocery store: "Every shelf stocked, every order shipped out the door," with weekly specials and same-day packing. The site went up in Sep 22 and has nothing to do with AI.
The same account has another app, Argon Flame, a casual game about sparks, updated Sep 21, with 0+ downloads. In Oct 20, dudufi.online went live with the one-line title "A little unexpected." In Oct 23, Dudufi was listed on Google Play. In Oct 27, its terms and privacy policy took effect.
By Nov 13 it ranked #11 in the Italy top grossing chart for this category, between AI Catch (7.56M installs) and Alight Motion (176M installs) (Google Play, 11-13).
What it is
The store description covers the features: browse templates with previews, pick a style you like, upload your own photo, and get a stylized image or an AI video. No prompts needed, and none can be written. One line in the description is blunt: "Browse template previews for free; purchase is required to generate videos and images." (Google Play, 11-12)
It makes money through "Stars" credits plus an auto-renewing membership, with one-off in-app purchases from $3.99 to $129.99. You can pay without signing up, and you link an email later (dudufi.online terms). A sibling app from the same playbook spells it out more clearly on its site: a template costs "80 Stars and up," standard video is 8 seconds, and membership unlocks 15-second videos and HD export.
The generation isn't done in-house. The privacy policy says photos go to BytePlus, ByteDance's model API service (Seedance for video, Seedream for images). Going by BytePlus's public pricing, a second of video costs roughly $0.07 to $0.23, so a 8-second video costs about $0.6 to $1.8 (estimate). Any template priced above that has a margin.
Identical ones, 4 more
Searching Google Play for "$3.99–$129.99", "Stars", and "template" turns up 5 apps. Their privacy policies match paragraph for paragraph, all naming BytePlus, AppsFlyer, and Meta, and all include a line about collecting SIM card state for ad conversion analysis and anti-fraud. Only the company names differ:
| App | Developers on Play | Company named in the privacy policy | Launched | Installs (11-13) | Top Grossing |
|---|---|---|---|---|---|
| Zivoo | NationPro | NATIONWIDE INSTALLATION SERVICES LLC | 09-09 | 219 | 11-13 charted, 198 |
| Bobofi | Bishomart | Syntrix AI Ltd | 09-30 | 3,752 | 181 → 104 |
| FableTicket | NLPInvest | Not stated | 10-13 | 2,433 | 168 → 104 |
| PoseAnew | FriendsJCC | Veyra AI LTD | 10-14 | 3,062 | 167 → 71; US 71 |
| Dudufi | JNH QuickShop | Jnhquicknshop LLC | 10-23 | 2,247 | 144 → 67; Italy 11 |
The top grossing chart is the Video Players & Editors category. The column with no country listed is the same chart shown with identical rankings across a dozen-plus countries, including Saudi Arabia, the UAE, Turkey, Poland, Mexico, and Vietnam (Google Play, 11-10 to 11-13).
The first line of each store listing also comes from the same hand: "A familiar photo. A whole new possibility." (Dudufi), "A photo of you. A story to step into." (FableTicket), "A little spark for the photos you love." (Bobofi), "A different outfit. An unexpected setting." (PoseAnew).
The accounts are farmed the same way. Besides Zivoo, each developer account also has an unused small game: Argon Flame, Onion Ring, Autumn Blaze, Water Sort Color Puzzle. Three of them share one package-name pattern: Greek letter plus word: com.xi.omicron.argonflame, org.psi.omega.onionring, com.alpha.autumnblaze (Google Play developer page, 11-12). The inference: use a small game to build up publishing history on the account, then launch the app that actually makes money.
The pace is one new reskin every two to three weeks, with old ones left up and paid acquisition continuing. Across the 5 apps, total worldwide installs are about 11.7k.
2 thousand installs, how it reached the top grossing chart
Three things came together.
Charge from the start. There is no free allowance; the first generation costs money. Big products like PixVerse and Kling give away credits to pull users in. Dudufi only wants people willing to pay, so installs are low but revenue still climbs.
Use payments to train ads. Privacy policy sections 1.4 and 1.5 state that purchase amount, currency, and order ID are shared with advertisers "for ad attribution and campaign measurement." The sibling app FableTicket's update notes say it outright: "Integrated AppsFlyer to attribute installs and measure purchase conversions." Once Meta knows who paid, it can look for more people like them. On Facebook there is a personal account with a garbled name, "Fduuduffjic Dudufi," that looks like an ad-buying account. The actual ad creative hasn't been found, so the channel is inferred, but all the evidence points to Meta.
Pick a low-bar chart. It isn't listed under Photo & Video or Art & Design, where AI portrait apps are packed. It sits under Video Players & Editors, where the top grossing list is mostly players, downloaders, and casting tools. 11-13 It ranks #67 in the overall chart, with LumaFusion (a pro editing app, 67k installs) on one side and Story Saver (13.9M installs) on the other. Its #11 in Italy also needs context: that day the Italy top grossing list for the category had only 16 apps.
The chart position itself brings little organic traffic. It shows that this team's paid acquisition is paying back, which is why it keeps launching new reskins.
Apart from paid acquisition, every other channel is zero. Searching TikTok for "dudufi" returns only a children's song. X had 60 posts in 0 days in the last stretch. No press coverage. The name is a meaningless coined word, and it doesn't look like it's getting store search traffic.
Who it's competing with
The photo-to-video template niche is already crowded, and Dudufi has no product advantage over others:
| Product | Launched | Scale (Google Play installs, 11-13) | Versus Dudufi |
|---|---|---|---|
| CapCut / Hypic (ByteDance) | 2020 / 2023 | 2.07B / 130M | Plenty of free templates; Dudufi uses ByteDance's models too |
| PixVerse | 2024-09 | 79.9M, +330k a week | Own models, free allowance, and free-form prompts |
| Kling AI (Kuaishou) | 2025-01 | 35.4M, +600k a week | Big-company in-house model, full feature set |
| AI Catch | 2025-08 | 7.56M | Same category, also template-based, driven by large-scale installs |
| Pollo AI | 2025-01 | 2.06M | Aggregates several models; #6 top grossing in the same category in Italy |
| ViralPet | 2026-10-01 | 24k | Pet videos only; #25 top grossing in the same category in the US |
| Clipsy AI, Video Me, FantasyPic, etc. | 2026-08 to October | 2–45k each | Another reskin matrix, in different categories |
The closest match to Dudufi is another matrix, and the one most worth studying is ViralPet: a new app of similar age, focused on one niche (pets), 10 times the scale, #25 top grossing in the US, while Dudufi ranks only #93 there.
If it were you
B: Worth trying, but what you'd learn is a paid acquisition playbook, not product features to copy. Nobody outside can see what this playbook returns.
One-line take: Dudufi shows that with a per-second video model API, a template app with a hard paywall plus Meta optimizing for purchases can reach the top grossing chart in two to three weeks, and one team can run 5 of them at once. Few people know this playbook yet, but the template-video niche is already crowded, and nobody knows what it earns in a month.
Who it fits: A 1–3 person team that has run Meta app ads, can send paid events back through AppsFlyer or similar tools, knows the cost per payer in each country, and has a test budget of a few thousand dollars. Ideally they can name 20 or more template concepts around a specific audience or scene: pets, weddings, babies, holidays, traditional dress from a particular country.
Who it doesn't fit: People who haven't run ads and expect organic store traffic (these apps get almost none; Dudufi's name doesn't even come up in search). People who want to build a branded AI video product that retains users (that road is taken by CapCut, PixVerse, and Kling). Anyone with a single Google Play account who can't afford a ban.
How to do it: Connect the backend directly to a per-second billed API like BytePlus Seedance or Seedream. Your job is three things: template design, paywall design, and acquisition. Don't build general styles; follow ViralPet and pick one concrete niche. Previews are free, the first generation is paid, and each price point should be at least three to five times the generation cost (a 8-second video costs about $0.6–1.8). Put a large credit pack in the top tier for high-spending users. Start with a category like Video Players & Editors, where the bar to chart is low.
How to acquire users: Meta app ads, sending purchase events back from day one so delivery optimizes for payers rather than installs. The creative is a before-and-after of an ordinary photo turned into a cinematic clip, one per template. Start with small budgets in cheap secondary markets (Dudufi's chart signals show up in Italy, the Gulf states, Turkey, Poland, Southeast Asia, and Latin America), and only scale creative that brings cost per payer below first-order revenue. Running several listings in parallel spreads the risk, but Google Play is strict about duplicate content and linked accounts, and batch bans are common.
How early you need to be: The sooner the better. Per-second video model APIs keep getting cheaper, so the barrier keeps dropping and more teams will follow. Launch and start small-budget tests before the end of December. If cost per payer still isn't below first-order revenue by Q1 of 2027, stop.
Signs you read it wrong: ① Over the next month, these 5 apps get pulled in batches, or their top grossing ranks fall out of the top 150 together, which means the playbook doesn't survive on Google Play or the acquisition doesn't pay back. ② Reviews fill up with one-star complaints like "paid but couldn't generate" or "charged a subscription," along with refunds, which means the hard paywall is buying refunds, not profit. ③ The reverse: if the team keeps launching a new app every two to three weeks and the older listings keep climbing, the return is positive and it's worth moving faster.
Data and sources
Installs and Top Grossing rank: Google Play public charts and detail pages, 2026-11-10 to 11-13; Dudufi detail page and developer page 11-12 read (English store, Italian store).
Links between the 5 apps: each one's website and privacy policy (dudufi.online, bobofi.store, poseanew.store, zivoo.tech, fableticket.site, read 11-12), Google Play developer pages. "Warm up accounts before listing" and "same team behind them" and "runs paid ads on Meta" are all inferences.
Costs: BytePlus ModelArk official pricing docs, plus third-party Seedance price roundups; cost per video is an estimate.
External channels: TikTok, X (last 60 days), web search. Found nothing about Dudufi or its sister apps. Not finding its ad creatives doesn't mean it isn't advertising.
Revenue: no public figures. A Top Grossing rank only shows it's taking money, not how much.
Research notes: competitor table, acquisition evidence, full research report
- Competitors
Name Where Launched Scale How it differs from this one Bobofi / FableTicket / PoseAnew / Zivoo (same matrix) Android 2026-09-09 to 10-14 Installs of 219–3,752 each; Top Grossing overall chart 71–198 Almost no difference: same backend, same privacy policy, same pricing. Only the brand name and developer account change. CapCut / Hypic (ByteDance) Big-company features 2020 / 2023 2.07B / 130M installs Big-tech tool with lots of free AI templates. Dudufi uses ByteDance's BytePlus models, charges only, and has no editing. PixVerse Android 2024-09 79.9M installs, +330k in 7 days Own model, free credits and free-form prompts. Dudufi only offers templates and gives no free credits. Kling AI (Kuaishou) Android 2025-01 35.4M installs Big-tech in-house model, full-featured, huge reach. Hailuo AI(MiniMax) Android 2025-01 28.1M installs Same as above AI Catch Android 2025-08 7.56M installs Template app in the same VIDEO_PLAYERS category, driven by huge install volume. Dudufi gets few installs but high spend per user. Glam AI Android 2024-02 6.24M installs Branded US template-video team, subscription model. Pollo AI Android 2025-01 2.06M installs Multi-model generation platform, #6 on the Italian VIDEO_PLAYERS Top Grossing chart. ViralPet Android 2026-10-01 24K installs, #25 on the US VIDEO_PLAYERS Top Grossing chart. Only does pet videos, a single use case. - Where users come from
- Inferred Meta paid ads: the privacy policy lists Meta as a recipient of paid-event data; Facebook has a garbled-name account used for warm-up, "Fduuduffjic Dudufi"; no creatives captured in the metaads library.
- Confirmed AppsFlyer attribution, purchase postbacks: Dudufi privacy policy sections 1.4 / 1.5; the sister app FableTicket's update notes say it integrates AppsFlyer to measure purchase conversions.
- Confirmed Multi-shell matrix scaling: 5 apps have privacy policies that match paragraph for paragraph (BytePlus, AppsFlyer, Meta, SIM detection). They're operated by 4 different LLCs / Ltds.
- Inferred Picking the VIDEO_PLAYERS category to grab Top Grossing slots: 2K installs climbed from the overall chart to #67 over 4 days, 144. Photography and Art & Design have higher bars.
- Inferred TikTok, X, Reddit, media, ASO: TikHub, x-radar, Jina and Brave all return nothing relevant. The name is a meaningless brand word.
Why Dudufi (com.dudufi.aigw) made it onto the charts
Bottom line: Dudufi itself isn't doing anything special. It's the 5th shell in a paid-acquisition matrix: same code, same privacy policy, same backend (BytePlus image and video generation APIs), same pricing ($3.99–$129.99 Stars credit packs plus a subscription). Only the name changes, along with an empty US LLC developer account. Between September and October it went live on Google Play as 5 apps. It reached the charts through three things: Meta ads (with AppsFlyer postbacks for paid events), a design where you pay before your first generation, and a low bar for the VIDEO_PLAYERS Top Grossing chart. Global installs are only 2,247, with no ratings, no reviews, and no press coverage, so almost nobody outside knows about it. The thing worth watching isn't this one app but the playbook: using ByteDance's own Seedance/Seedream APIs to mass-produce "photo plus template becomes video" apps at low cost.
---
1. What it is
- An AI app that turns photos into styled images or a 8–15 second short video using templates (cyberpunk, neon, outfit swaps, scene changes and the like). You can't write free-form prompts. The store description says "Template previews are free, generation requires payment" (Google Play detail page, Italian store, read 2026-11-12).
- Billing: Stars virtual credits plus an auto-renewing membership, in-app purchase range $3.99–$129.99 (googleplay library app_snap, 11-10 to 11-13 all show this range). The website says "Creating uses Stars" (dudufi.online). The sister Bobofi site says "STARTING AT 80 Stars", with memberships giving weekly Stars, 15-second videos and HD export. FableTicket says "standard 8 seconds, members unlock 15 seconds".
- Accounts are anonymous: you can pay without signing up, and add an email later (dudufi.online/terms, effective 2026-10-27).
- Listed under Video Players & Editors, not Photography or Art & Design. Content rating PEGI 3.
- Developer: Play shows JNH QuickShop, legal entity Jnhquicknshop LLC, address 182 E 1975 N Centerville, UT, contact email [email protected] (Play detail page). jnhquickshop.buzz is a template-style "online everyday goods store" website with nothing to do with AI. A same-named eBay shop, jnh_quicknshop, also turns up (Jina search). It looks like a shell company, or a rented account, used to get through Play's organization verification.
- Release dates: store launch 2026-10-23; website launch 2026-10-20; terms and privacy policy effective 10-27; latest version updated 1.0.1,11-07.
2. Data: how big it really is
| Date | Cumulative global installs | VIDEO_PLAYERS Top Grossing chart (overall chart*) | Other |
|---|---|---|---|
| 11-10 | 1,615 | 143–148 (14 store) | Charted on day one |
| 11-11 | 1,615 | 101–119 | |
| 11-12 | 2,247(+632) | 82–92 | |
| 11-13 | 2,247 | 67–82 (14 store) | Italy 11 (this chart only had 16 apps); US 93 (127 apps in total) |
Source: googleplay library chart_rank / app_snap, 2026-11-10 to 11-13.
\* On this day the VIDEO_PLAYERS Top Grossing chart for ae / ar / mx / my / ng / nl / ph / pl / sa / sg / th / tr / vn / za was also exactly 200 places, with nearly identical rankings, so it's the same overall chart. The card says "charted in 15 countries", but in reality there's one overall chart plus two real country charts, Italy and the US.
- Over those four days this overall chart climbed from 144 to 67, which suggests paid revenue is growing. 11-13 on either side of it sit LumaFusion (the professional editor with 6.67M installs) and Story Saver (6.81B installs). 2 thousand installs ranking here means each install brings in a lot of revenue. A rough ARPU estimate puts it at possibly dozens of times the category average. That's an inference, not verified.
- Treat the Italian #11 with caution: the Italian category chart only captured 16 apps, so the bar is low. Still, on that day it did rank ahead of AI Catch (7.56M installs) and Alight Motion (176M installs) (googleplay library v_chart, it, 11-13).
- No ratings, no reviews (ratings empty in app_snap; recent_reviews empty on the card). AppstoreSpy has no history for it (0 rows in app_day_ext).
- Roughly 300 new installs per day (11-10 to 11-12 gave +632). Very small scale, like a test budget during paid acquisition.
3. Competitors
3.1 Sister apps in the same matrix (the most important finding)
The googleplay library turns up 5 apps. They all have the exact same in-app purchase range, $3.99–$129.99, all sit in the VIDEO_PLAYERS category, and their descriptions mention "Stars" and "template". Their websites all use a single brand word plus a cheap suffix, with a /privacy path. I read the privacy policies with Jina, and the body text matches paragraph for paragraph: all mention BytePlus, AppsFlyer and Meta, and all include the paragraph on collecting SIM card status for ad attribution and anti-fraud analysis. Only the operating entity name differs:
| App | Bundle ID | Play developer name | Operating entity named in the privacy policy | Launched | Installs (11-13) | top grossing best rank |
|---|---|---|---|---|---|---|
| Zivoo | tech.zivoo.app | NationPro | NATIONWIDE INSTALLATION SERVICES LLC | 09-09 | 219 | 198 (Poland overall chart, only charted from 11-13) |
| Bobofi | bobofi.store.ai | Bishomart | Syntrix AI Ltd | 09-30 | 3,752 | Overall chart 181 → 104 |
| FableTicket | com.fableticket.app | NLPInvest | (website FAQ says 8-second / 15-second member videos) | 10-13 | 2,433 | Overall chart 168 → 104 |
| PoseAnew | com.poseanew.gigv | FriendsJCC | Veyra AI LTD | 10-14 | 3,062 | Overall chart 167 → 71; real US chart #71 |
| Dudufi | com.dudufi.aigw | JNH QuickShop | Jnhquicknshop LLC | 10-23 | 2,247 | Overall chart 144 → 67; Italy #11 |
A few details:
- The store copy comes from the same writer. "A familiar photo. A whole new possibility." (Dudufi), "A photo of you. A story to step into." (FableTicket), "A little spark for the photos you love." (Bobofi), "A different outfit. An unexpected setting." (PoseAnew).
- FableTicket's update notes state outright: "Integrated AppsFlyer to attribute app installs and measure purchase conversions". That is effectively an admission that it runs paid ads.
- The package suffixes are messy (.aigw, .gigv), and the websites launched in batches, with FableTicket and Bobofi at 9-24 and Dudufi at 10-20. The cadence is roughly a new shell every 2–3 weeks, while the older shells keep spending.
- Each shell's curve looks much the same: after charting, a few hundred installs a day, with Top Grossing rank climbing steadily. PoseAnew went from 280 to 3,062 in a week; Bobofi went from 779 to 3,752 in two weeks (googleplay library app_day).
So Dudufi's first competitors are its own siblings. The playbook is one team running A/B tests across different accounts, which also spreads out the risk of bans.
3.2 External competitors (doing the same thing)
| Name | Where | Launched | Scale (googleplay library, 11-13) | How they differ from Dudufi |
|---|---|---|---|---|
| CapCut / Hypic (ByteDance) | Android, iOS | 2020 / 2023 | 2.07B / 130M installs | Big-tech tool with lots of free AI templates. Dudufi uses ByteDance's BytePlus models. |
| PixVerse | Android, iOS, web | 2024-09 | 79.9M, 7 days +330k | Own model, templates plus free-form prompts, with free credits. |
| Kling AI (Kuaishou) | Android, iOS, web | 2025-01 | 35.4M, 7 days +600k | Big-company in-house model, full feature set |
| Hailuo AI(MiniMax) | Android, iOS, web | 2025-01 | 28.1M | Same as above |
| DreamFace | Android, iOS | 2023-03 | 11.2M | Long-running face-swap and photo-animation template app, subscription model. |
| AI Catch | Android | 2025-08 | 7.56M | Similar template app in the VIDEO_PLAYERS category, far larger in scale. |
| Glam AI | Android, iOS | 2024-02 | 6.24M | Branded US template-video team, #13 on the Italian Photography Top Grossing chart. |
| Pollo AI | Android, web | 2025-01 | 2.06M | Multi-model aggregator, #6 on the Italian VIDEO_PLAYERS Top Grossing chart. |
| ViralPet | Android | 2026-10-01 | 24k | Only pet video templates, VIDEO_PLAYERS US Top Grossing 25 |
| Clipsy AI / Video Me / FantasyPic / ReeLora / Framegen | Android | 2026-08 to October | 2–45k each | Another matrix (privacy policies hosted at file.<brand>.xyz/…/privacy.html), parked in Art & Design / Entertainment, driving heavy installs |
Source: googleplay library app_profile.
Against external competitors, Dudufi has no product differentiation: the model is bought (BytePlus), and the templates are generic. Only two things are actually different:
- A hard paywall from day one. No free quota. Preview is free, generation is paid, credit packs up to $129.99. PixVerse and Kling use free quota to pull in users first; Dudufi only wants people who will pay.
- Filed under VIDEO_PLAYERS. Photography and Art & Design are packed with AI photo apps, so the grossing bar is high. The VIDEO_PLAYERS grossing chart is mostly players, downloaders and casting tools, and 2 thousand installs is enough to rank in the top 70 of the overall chart.
4. Growth tactics: where users come from
| Channel | Evidence | Confirmed? |
|---|---|---|
| Meta paid acquisition (FB / IG) | Section 2 of the privacy policy lists "Meta and Other Advertising Partners" as recipients of paid-event data. There is a Facebook personal account with a garbled name, "Fduuduffjic Dudufi" (id 61595973546433), that looks like an account warmed up for ad delivery. Our metaads library doesn't show it (the library only tracks listed advertisers and first-screen keywords). | Partly confirmed (intent and infrastructure present, no specific creatives captured) |
| AppsFlyer attribution, paid-event postbacks | Dudufi privacy policy, sections 1.4 / 1.5: purchase amount, currency and order ID are "used for ad attribution and campaign measurement". The sibling FableTicket's update notes say it integrates AppsFlyer. | Confirmed |
| SIM card and regional analytics | Privacy policy section 1.6: collects SIM status, and combined with channel attribution, payments and retention to analyze conversion and retention by audience segment, plus anti-fraud. This is a common practice among China-based going-global paid acquisition teams. | Confirmed (the text exists) |
| Matrix spam (many accounts, many shells) | 5 shells, see 3.1 | Confirmed |
| Category choice to grab chart slots | After being filed under VIDEO_PLAYERS, the overall grossing chart climbed from 144 to 67 over 4 days | Confirmed (the effect); whether it was deliberate is inference |
| TikTok / creators | Searching "dudufi" on TikHub returns only unrelated nursery-rhyme videos ("dudi dudi dam dam"); the tiktokads library has no mapping. | Nothing found |
| X / Reddit / press | x-radar search for dudufi / poseanew / bobofi / fableticket returned 60 posts over 0 days. Jina and Brave web search turned up nothing beyond the brand's own site and AppBrain listings. No press coverage. | Nothing found |
| Store search (ASO) | The name is a meaningless brand word, and the titles contain no "AI video" keywords, so it doesn't look like it's pulling in users through search. | Inference: not a major channel |
| Google Ads | In the googleads library's watch table, PoseAnew's status is unchecked; Dudufi is not on the list. | Not verified |
Where the money comes back from: Users see "photo to cinematic movie" creatives on Meta, tap in, watch the template preview for free, then pay on their first generation (Stars pack or weekly membership). AppsFlyer sends the payment back to Meta, and Meta optimizes delivery toward people likely to pay. On the cost side: BytePlus Seedance's official pricing is roughly $0.07–0.23 per second (480p–720p; from BytePlus docs and the framesurfer and camclo3d pricing roundups). A 8-second video costs about $0.6–1.8. A template priced 80 Stars only has margin if it is priced above that figure; the $129.99 big pack is for high-paying users. The target markets are secondary markets like the Gulf, Turkey, Italy, Southeast Asia and Latin America, where paid acquisition is cheaper than in the US. Which countries they actually spend in is inference: paid-chart signals only show up in Italy and the overall-chart countries, and the US only ranks 93.
5. Why it's taking off (at its current scale)
- Seedance-style model APIs have become cheap and open (timing). BytePlus ModelArk bills per second, so a small team doesn't need to build its own model and can wrap a shell and ship in two or three weeks. The privacy policy states generation is done by BytePlus.
- Matrix replication pushes marginal cost to near zero (tactic). The same code, copy templates and store page templates, with a new shell and developer account every 2–3 weeks. If one shell gets banned or stops performing, the whole operation isn't hurt.
- Paywall up front plus Meta value optimization (tactic). No free quota; the first generation is charged. Then AppsFlyer sends paid events back so Meta finds people who will pay. That's why installs are low but grossing ranks still climb.
- Category arbitrage (tactic). The VIDEO_PLAYERS grossing bar is low: 2 thousand installs gets you into the top 70 overall and the top 11 in Italy. The chart position itself brings very little organic traffic to a paid-acquisition product, though; its main effect is that our recall picked it up.
6. Differentiation
No product differentiation. The templates are generic, the model is bought in, and the UI matches its sibling apps. Everything that looks "different" is operational: the hard paywall, high-priced credit packs, category selection, the multi-shell matrix. Those are why it can chart, but they belong to this team, not to the Dudufi product. Any going-global paid-acquisition team can copy the same playbook, and several matrices are already running it (see the last line of 3.2).
7. Visibility
Almost nobody knows about it yet. No ratings, no reviews, 0 posts on X, no press coverage, revenue not disclosed. But the "photo template AI video" category is already crowded, with big players (ByteDance, Kuaishou, MiniMax) and dozens of matrix shells in it. The category isn't short on visibility. What's missing is for this low-install, high-ARPU approach to get noticed.
8. What others can borrow
- The model is taken directly from a per-second billed API like BytePlus / Seedance, and the product only handles template design and payment design.
- No free quota. Preview is free, generation is paid. This filters users down to paying users, and the paid events are then used to train Meta delivery.
- Pick the category with the lowest grossing bar (VIDEO_PLAYERS) so a small scale is enough to chart.
- Parallel multi-shell testing: same backend, different names and accounts; add budget to whichever shell has the best ROI. The risk is Google Play's policies on duplicate content and linked accounts, and mass bans are common.
- A risk-control playbook common among Chinese teams: anonymous accounts to lower the paywall barrier, and SIM card status plus root detection for anti-fraud and audience segmentation.
9. Still unclear
- Actual revenue. We don't know what 67 overall-chart rank corresponds to in daily gross; it hasn't been calibrated. We only captured 16 Italy-chart apps, so the weight of the ranking is limited.
- Which company is behind it. The operator uses 4 differently named LLCs / Ltds. Judging by the BytePlus use, SIM detection and privacy policy wording, it's most likely a China-based going-global team, but this is unverified. Whether RolePlai, which appears as Veyra AI on iOS, is the same company is also unverified.
- Paid acquisition platforms and creatives. Only indirect evidence exists (privacy policy and warmed-up Facebook account). We didn't capture specific ads in the Meta ad library. Whether they also spend on Google UAC or TikTok is unclear.
- The recall card mentions "France new-release free chart #4", but that chart isn't in our library, so we can't verify it.
- Retention and refunds. No reviews, so we can't tell whether paying users are satisfied. Hard-paywall template apps like this usually collect bad reviews and refunds; we can check once reviews come in.
- How long it lasts: sibling Bobofi launched 6 weeks ago and is still climbing. The matrix shows no sign of decline or bans yet.
Sources
- googleplay library (agent-im server, DuckDB): chart_rank, app_snap, app_day, app_profile, app, 2026-10-29 to 11-13
- metaads / tiktokads / googleads / appstore libraries: searched Dudufi and sibling apps, no results (PoseAnew's status in Google Ads is unchecked)
- www.dudufi.online/, /terms, /privacy (read via Jina, 2026-11-12)
- www.poseanew.store/privacy, www.bobofi.store/privacy, www.zivoo.tech/privacy, www.fableticket.site, www.bobofi.store
- Google Play listing play.google.com/store/apps/details?id=com.dudufi.aigw (gl=it)
- jnhquickshop.buzz/; Facebook personal account www.facebook.com/p/Fduuduffjic-Dudufi-61595973546433/
- BytePlus ModelArk pricing docs.byteplus.com/en/docs/ModelArk/1544106; framesurfer.com/blogs/seedance-2-0-pricing; camclo3d.com/blog/seedance-pricing
- x-radar (60 days, 0 posts), TikHub TikTok search, Jina / Brave web search