1.2M installs in three months, then a loss of $1,339 in September: the books of an AI photo template app
What it is Upload a photo, pick a template, and in seconds it turns you into anime art, an aged version of yourself, a muscled version, a couple-in-the-rain picture, or a short video. An AI effects app.
Why it blew up The product isn't new, the distribution is aggressive: a studio with paid-acquisition experience ran multiple accounts, bought installs in cheap markets like India, Nigeria and the Gulf, and paired that with store pages tuned per country and pre-seeded five-star reviews.
CHow you could follow it Stop building the everything-app. Pick one scene in one country (say Indian weddings or festival greeting cards) and go deep. Charge per output, run before-and-after short videos on a small paid budget, and add the social sharing channel it's missing.
- Data (updated daily) · first launch 2026-07-21 · 11-07 installs 1,200,098
Android · Total installs worldwide Data 2026-10-29 → 2026-11-07 (10 days, 10 observations). Google's detail page only refreshes install counts every other day, so a flat day doesn't mean nobody installed.Android · Ratings count (cumulative) Data 2026-10-29 → 2026-11-07 (10 days, 10 observations)
Picsgen is an app from a Pakistani studio, published through sock-puppet accounts: upload a selfie, switch styles with one tap. It launched at the end of July, and by November it had 6 installs on Google Play, 1.2M of them, mostly from India. It shows two things. In emerging markets, this kind of app can buy a million installs within weeks. And buying them doesn't guarantee profit: its source code is already for sale, and the books show a loss in September.
A five-star review with 159 people marking it helpful
Opening Picsgen's India store page (11-06), the title reads "PicsGen - AI Image & Video", rated 4.2 stars. The top reviews pinned to the front: one from Oct 12, "This is my first time downloading, installing and using this app, the developer did an amazing job!!!", found helpful by 152 people; another from Sep 8, signed nabeel abid, whose entire text is "amazing quality result and templates", found helpful by 159 people. The first real Indian user review is from Satyam Bhoi (10-16): "This app is unfinished, the ads don't even show 😡".
The same app's US store rating is 5.0,831 ratings, all one to four stars 0 (Google Play rating distribution, 11-06).
Scrolling further, the developer website field says terafort.com, and the support email is [email protected]. Opening the privacy policy link gives the privacy policy of a note-taking app the same account published in June, reused as is.
What it is
A templated AI photo and video effects app: upload a photo, pick a template (anime style, aging via Future Me, muscle transformation via Body Builder, old photo restoration, background removal, hair color try-on, image-to-video), and get an output in seconds. The India listing adds local touches: "Couple in Rain", "Love Moments", birthday videos, festival greeting cards. The selling point is one line: no prompt writing needed.
It makes money two ways: in-app purchases of $1.99–$69.99 each (credits or subscription), plus ads. The backend most likely calls third-party image and video generation APIs (inferred), so there's no real technical barrier.
Who made it
The developer account is called "AppsVariety Digital", behind which is Terafort, a company with an office in Lahore, Pakistan, that makes mobile games and tools. Its website claims "5 games, 50M+ downloads", and its roadmap lists a "Gen-AI Social Media App". Matching by email and domain, Terafort has at least 7 Google Play developer accounts. "gpc8" looks like "developer console #8" (inferred).
It's also copying itself: in Aug 28, another Terafort account, Nubivio Apps, published "AI Photo: Magic Editor pro", with the contact email [email protected]. "ASO + AppsVariety", a team dedicated to pumping installs. This copy has 11-06 installs, 108k of them.
Where the traffic comes from
Mostly bought (inferred, with fairly high confidence). A third-party SDK scan (AppGoblin, August) shows it integrates Adjust attribution plus ad networks including Facebook, AppLovin, Unity and Google, 11 in total. An app relying only on organic traffic wouldn't need paid attribution. The countries where it charts tell the story too: India, Nigeria, the UAE, Saudi Arabia, South Africa, Brazil, Malaysia. All cheap-acquisition markets, with none of the US, UK, Germany or Japan. I couldn't find its ad creatives, and how much it spent isn't visible from outside.
Store search is finely tuned. The India and US stores use two different titles, and the descriptions cram in long-tail keywords like AI photo enhancer, background remove, collage maker, couple in rain, hair colour, double exposure, and birthday videos. Updates come almost weekly (11-03 to 1.33).
Seeded reviews drive conversion. All 831 US ratings are five stars, while India has only 4.2. The pinned reviews are written and signed in ways that don't match the target users. This is a common cold-start tactic, but it also means the star rating says little about real word of mouth.
No content channel. Searching "picsgen" on TikTok turns up no related videos. On X, zero mentions in 90 days, and no press coverage (11-06).
The result is this curve (Google Play installs):
| Date | Total installs | India Photo & Video, Top Free |
|---|---|---|
| 07-21 | Launched | — |
| 10-29 | 993k | #16 |
| 10-30 | 993k | #14 |
| 11-02 | 1.08M | #21 |
| 11-05 | 1.2M | #35 |
| 11-06 | 1.2M | #40 |
It added about 208k in a week, but its free chart rank kept falling. Its growth comes from sustained ad spend, not organic word of mouth. 11-06 It entered India's Photo & Video Top Grossing chart for the first time, at #34; the grossing bar in that category in India is low.
Money: source code for sale and an ugly ledger
A source code marketplace, AppWill, lists "PicsGen: AI Photo Editor": full Kotlin source code, code only without the developer account, priced at $7,455. The listing gives monthly net revenue of $305 for July, $2,003 for August, and -$1,339 for September, a three-month total of $968. The name, launch month and tech stack match, so this is most likely this app or the same codebase (inferred, since the listing doesn't give a package name).
A separate third-party tool estimates its revenue at more than $50k a month, an order of magnitude off the listing. The listing's "net revenue" may already be net of ad spend, and gross revenue isn't visible from outside. Either way, it's not an easy business: the users it bought barely pay back, so the studio is also selling the code. Ads, in-app purchases, and selling the source code: three revenue streams, all needed.
Who it's competing with
Almost everyone. Apps with the same features flood Google Play in the same period (11-06 installs):
| App | Launched | Total installs | past 7 days | Versus Picsgen |
|---|---|---|---|---|
| Refoto: AI Image & Video Maker | 2025-12 | 24.6M | +590k | Also focused on India, 20 times the size |
| Muselab AI | 2026-04 | 5.36M | +920k | Same kind, charting in 30 countries |
| AI Face Aging | 2026-02 | 7.13M | +570k | Only the aging feature |
| AI Photo Beauty Editor Studio | 2026-03 | 2.97M | +810k | Same kind, 30 countries |
| Art Magic and a batch of similarly named apps | 2026-05~09 | 100–4M | +25~500k | Almost identical |
| Remini / Hypic / PixVerse | 2019–2024 | 0.8–620M | +50~800k | Established players, big companies, in-house models |
| Gemini and ChatGPT image features | From ¥2025 | Hundreds of millions of users | — | Free, but you need to know how to write prompts |
Ranked by new weekly installs among AI photo apps, Picsgen's 208k doesn't make the top 30 (#30 is about 240k). It isn't the fastest runner in the category, just a mid-pack one.
What it got right and what it didn't
What it got right was all distribution: a studio with paid-acquisition experience, multiple developer accounts, store pages and templates tuned per country, reviews seeded in advance, and cheap countries bought first. None of the product itself is something others don't have.
What it didn't do is keep users. Indian negative reviews cluster around "it doesn't generate something new, it just gives me the existing templates" and "the features aren't finished". Its rank fell from 14 to 40 within a week, and its September books are in the red.
If it were you
C. You can learn a full playbook for buying installs in emerging markets, but copying an AI template app will most likely mean losing money on paid installs in a crowded category.
Who it suits: small teams that have already bought installs in India, Nigeria or the Middle East, who know their per-install cost and payback period, and who have reusable code and several developer accounts. For you, this is a standard playbook, not a new opportunity.
Who it doesn't suit: anyone who hasn't run ads and hopes organic traffic or TikTok will carry them, or anyone aiming to win on "more templates", since a dozen apps with identical feature lists launched in the same period.
How to do it: if you do build one, don't build the everything-app. Pick one local scene and go deep, like Indian weddings or Diwali greeting cards, or a country's ID photos or religious festivals. Write the templates and copy for that country's people, so it's obvious at a glance that this was made for them. Sell per output, priced at what locals can pay (₹50–200 per output), and let people generate one image free first. Run generation through a third-party API, and calculate the cost per image before you start.
How to acquire users: test creatives on Meta and AppLovin with a small budget in India and Nigeria. The creatives should be vertical before-and-after short videos, one per template. Write store page titles and descriptions per country, packed with long-tail keywords. In the first two weeks after launch, get real users to leave reviews. One way Picsgen differs: have the output carry a watermark and download link, and try to get it shared on WhatsApp and Instagram. Picsgen has no content channel at all, and that's the gap you can fill.
When to move: festivals are the peak for this kind of app. India's wedding season runs from November to February of the following year. To catch it, launch and test before December. If you miss it, wait for the next festival.
What would mean you got it wrong: ① if Picsgen or a similar app like Refoto holds a top-20 spot on India's Photo & Video Top Grossing chart for over a month, the business pays better than it looks, and the math needs redoing; ② if this source code stays unsold on AppWill for a long time or drops sharply in price, even the people who built it don't want it, so your judgment was right; ③ if Google or Meta offers the same one-tap templates for free directly inside WhatsApp or Instagram, then even the middle tier is gone.
In one line: in the AI template category, generating an image is no longer worth much. What's worth something is which country you buy a paying user in, and how much it costs you to do it.
Research notes: competitor table, acquisition evidence, full research report
- Competitors
Name Where Launched Scale How it differs from this one Picsart AI Photo Editor Android 2011 1.45B installs, +1.77M a week Full editor + community + brand. Picsgen only has template generation. Remini Android 2019 620M installs, +830k weekly Established AI retouching/portrait app with global subscriptions and paid acquisition. Picsgen only runs in emerging markets. Hypic (ByteDance) Android 2023 130M installs, +760k weekly Big-company template ecosystem plus TikTok traffic PixVerse Android 2024-09 79.6M installs, +520k weekly In-house video model, global Top Grossing Hailuo AI(MiniMax) Android 2025-01 27.8M installs, +530k weekly Image-to-video from an in-house model Refoto: AI Image & Video Maker Android 2025-12 24.6M installs, +590k weekly Template clone also focused on India, 20 times the size Muselab AI Android 2026-04 5.36M installs, +920k weekly Similar clone, charting in 30 countries with more aggressive buying AI Photo Beauty Editor Studio(ATQ - AI Lab) Android 2026-03 2.97M installs, +810k weekly Similar clone, covering 30 countries AI Face Aging(Willson Studio) Android 2026-02 7.13M installs, +570k weekly Single-feature aging, matches Picsgen's Future Me Google Gemini / Nano Banana, ChatGPT image generation Big-company features 2025 Hundreds of millions of users Free but you write the prompts; Picsgen sells one-tap templates - Where users come from
- Inferred Paid acquisition (Meta/AppLovin/Unity/Google): AppGoblin detected the Adjust attribution SDK plus 11 ad network SDKs; the charting countries are all low-CPI paid-acquisition markets like India, Nigeria, the Gulf, South Africa and Brazil; AppWill lists it for sale with September net revenue of -$1,339. Not in the Meta ad library, no creatives seen
- Confirmed Store search / ASO + category charts: the India store page is retitled PicsGen - AI Image & Video, with the description stuffed with localized long-tail keywords; Google results carry utm_term=picsgen+ai+photo+editor; in India's Photography Top Free, 10-29~11-01 holds the top 20
- Confirmed Seeded reviews: US-store rating histogram is 100% five-star across 831 ratings (googleplay.app_snap); India store is 4.2 stars; the top-liked positive review is from nabeel abid (158 helpful votes)
- Confirmed Sock puppet account matrix: developer email [email protected]; Terafort owns at least 7 developer accounts; a same-family clone, AI Photo: Magic Editor pro; the privacy policy reuses a blogspot page from a note-taking app
- Inferred TikTok / X / press: TikHub search for picsgen returns no relevant videos; x-radar shows 0 posts over 90 days; no press coverage found
Why Picsgen – AI Photo Editor (com.picsgen.ai.art.photo.editor) worked
Research date 2026-11-06. Sources: our googleplay database (app / app_snap / chart_rank / app_profile, from 2026-10-29), the metaads database, the Google Play store page (India), AppBrain, AppGoblin, apppage.net, the AppWill listing, the Terafort website, plus Jina / Exa / TikHub / x-radar searches.
Bottom line
It's not a product-innovation story. It's Terafort, a tools and games studio in Lahore, Pakistan, running a templated AI photo/video effects wrapper app under a faceless developer account called "AppsVariety Digital." It was pushed with paid user acquisition (wired to Adjust attribution and 11 ad SDKs including Facebook, AppLovin, and Unity), store search and category charts in emerging markets like India, and seeded five-star reviews (100% five-star in the US store). Installs were driven to 1.2M, but revenue was thin: the source code of the same-named app is listed for sale on AppWill, and the listing's financials show 7–September net revenue totaling about $968, with September at -$1,339 (they spent more on acquisition than they earned).
1. Basic data
| items | Value | Source |
|---|---|---|
| Launched | 2026-07-21 | googleplay.app(released_store) |
| Total installs | 1,200,098(11-06) | googleplay.app_snap |
| New in the last 7 days | 207,526 (10-29 → 11-05, about 30k/day) | googleplay.app_profile |
| past 30 days | About 510k (about 17k/day on average) | AppBrain |
| Over about 10-31 and 4 weeks | +454k, +102k that week | AppGoblin |
| Rating | US store 5.0, rating histogram [0,0,0,0,831] — all 831 ratings are five stars | googleplay.app_snap(detail_cc=us) |
| India store | 4.2 stars, 828 reviews | Play store page (hl=en_IN) |
| In-app purchases | $1.99–$69.99 per item; includes ads | googleplay.app_snap |
| Version | 1.32 → 1.33 (11-03), iterated to v1.27+ by October, nearly a weekly release | app_snap, apppage.net |
| Countries charted | 7–8 of them: in / ng / ae / sa / za / br / my (10-29 also th) | chart_rank |
Rank trend (googleplay.chart_rank, Photography category Top Free)
- India: 10-29 #16 → 10-30 #14 → 11-02 #21 → 11-05 #35 → 11-06 #40; 11-06 first entered India's Photography Top Grossing at #34.
- Nigeria 54 → 81; UAE 81–102; Saudi Arabia 98–145; South Africa 89–120; Brazil 137–162; Malaysia 137–178.
- Takeaway: the main battlefield is India, followed by Nigeria, the Gulf and South Africa; no developed market at all (US, UK, Germany, Japan none charting). Free-chart rank is sliding, and install growth comes from sustained spend rather than organic breakout.
2. Who made it: Terafort's sock-puppet account matrix
- The developer "AppsVariety Digital" lists
https://www.terafort.com/as its developer website, with email[email protected](googleplay.app). "gpc8" looks a lot like "Google Play Console no. 8". - The privacy policy link is
appsvarietydigital.blogspot.com/2026/06/alpnotes-note-taking-app.html, a note-taking app's privacy policy reused directly, a typical trace of bulk listing. - Terafort: describes itself as a startup that "develops and publishes mobile apps and games", with offices in Lahore (Facebook page, rozee.pk job listings). It also has Terafort US LLC (App Store), and Tracxn records an Austin company founded in 2021. Its website says "5 games, 50M+ downloads" and lists a "Gen-AI Social Media App" under Upcoming (terafort.com).
- Developer accounts under Terafort in our database (by email / domain): at least 6, including Terafort IEG, Funverse IEG, TF Incredible Apps, Nubivio Apps, Crownforge Games, EVERON Playworks, and AppsVariety Digital (googleplay.app, emails support@ / games@ / gpc7@ / gpc8@ terafort.com, asoappsvariety@ / incredibleapps@ terafort.org).
- Same-family clone: Nubivio Apps' "AI Photo: Magic Editor pro" (email [email protected], 2026-08-28 listed, 108k installs, on the Nigeria Photography Top Free chart). The same company swapped accounts and ran the same thing again. The email prefix "asoappsvariety" shows AppsVariety is an ASO / bulk-install project.
3. Growth tactics (evidence)
- Paid acquisition (likely the main driver, evidence is indirect): AppGoblin detected the Adjust attribution SDK and 11 ad networks including Facebook / AppLovin / Unity / Google, plus AdMob mediation (AppGoblin, SDK scan 2026-08-23). Apps that rely only on content or organic traffic don't need paid attribution. The AppWill listing shows September net revenue of -$1,339, consistent with spending more on acquisition than it returned. Our Meta ad library didn't catch it (we only track Pages surfaced by keyword searches, so not finding it doesn't mean it wasn't advertised). CPIs in India, Nigeria, the Gulf, South Africa and Brazil are very low, so 300k-per-week installs cost little.
- Store search / ASO: the India store page is retitled "PicsGen - AI Image & Video" (the US listing is still "Picsgen – AI Photo Editor"), showing a custom store page per country. The description is stuffed with keywords (AI photo enhancer, background remove, collage maker, couple in rain, hair colour, double exposure, birthday videos, festival photos...). Google store links in Jina search results carry the organic search parameter
utm_term=picsgen+ai+photo+editor. apppage.net gives it an ASO score of 86. - Seeded reviews: 100% five-star across 831 US-store ratings (app_snap histogram), while the India store sits at 4.2 stars. The two top-ranked positive reviews on the India store page received 151 and 158 "helpful" votes. One is from "nabeel abid" (a common Pakistani name) and says only "amazing quality result and templates". These look like incentivized or self-posted reviews. That's a legitimate cold-start tactic, but its star rating doesn't reflect real sentiment.
- Category chart placement: it held the top 20 of India's Photography Top Free for at least a week (10-29~11-01), and that placement itself brings organic traffic.
- Content / creators: TikTok searches for "picsgen" and "picsgen ai" return no relevant videos (TikHub, 2026-11-06). On X (x-radar, 90 days) there are 0 posts. No press coverage found. No content channel.
- Free promotions / celebrity / platform events: none found.
4. Why users come and why they pay
- Why it worked: demand in emerging markets for "upload a selfie → one tap turns it into a trending style / couple photo / birthday video" was trained by Remini, Hypic, PixVerse and the Nano Banana trend wave. Picsgen just bought traffic on that demand and caught the search terms. The description specifically localizes for India (Couple in Rain, Love Moments, Celebrations / festivals).
- How it makes money: in-app purchases at $1.99–$69.99 (per item, credits / subscriptions), plus ad monetization (AppGoblin estimates IAP 76% / ads 24%; its "$50K+/month" estimate is more than an order of magnitude off the AppWill listing, so the listing is more credible). Breaking into India's Photography Top Grossing at #34 (11-06 first entered the chart) has a low bar.
- Retention: negative reviews cluster around "doesn't generate new ones, only gives ready-made outputs" and "features incomplete, ads don't show" (Play India store, AppBrain reviews, 2026-10). No retention data available; the AppWill page claims "high retention", which we haven't verified.
5. Money and visibility
- AppWill lists "PicsGen: AI Photo Editor" priced at $7,455, "$968 net profit over the last 3 months", with monthly net revenue: 2026-07 $305, 2026-08 $2,003, 2026-09 -$1,339. Full Kotlin source, "source code only, no developer account included" (appwill.co/product/picsgen-ai-photo-editor/, scraped via Exa). The name, launch month (July), and tech stack (Terafort's job listings say Kotlin + Compose) all line up, so the source for this app is very likely what's for sale (or the same seller's codebase), though the listing doesn't give the package name, so we haven't verified it to 100% confidence.
- Visibility: almost nobody knows about it. No founder posts about revenue, no press, zero discussion on X and TikTok. Only automated aggregator pages like AppBrain, AppGoblin, Softonic and Uptodown. The only "public" revenue figure is the AppWill listing, and the numbers aren't flattering.
6. Competitors
| Competitors | Where | Launched | Scale (googleplay library, 2026-11-06) | How it differs from Picsgen |
|---|---|---|---|---|
| Picsart AI Photo Editor | Android / iOS / web | 2011 | 1.45B installs, +1.77M a week | Full editor + community + brand; Picsgen is only template generation |
| Remini(Bending Spoons) | Android/iOS | 2019 | 620M, +830k weekly | Established retouching / AI portrait app with mature subscriptions and global paid acquisition |
| Hypic (ByteDance) | Android/iOS | 2023 | 130M, +760k weekly | Big-company templates + TikTok ecosystem traffic |
| PixVerse | Android / iOS / web | 2024 | 79.6M, +520k weekly | In-house video model, global Top Grossing |
| Hailuo AI(MiniMax) | Android / iOS / web | 2025-01 | 27.8M, +530k weekly | Image-to-video from an in-house model |
| Refoto: AI Image & Video Maker(mAst App 2026) | Android | 2025-12 | 24.6M, +590k weekly | Same kind of template wrapper, also targeting India, 20 times the size |
| Muselab AI(Shotcut Video Workshop) | Android | 2026-04 | 5.36M, +920k weekly | Similar wrapper, charting in 30 countries |
| AI Photo Beauty Editor Studio(ATQ - AI Lab) | Android | 2026-03 | 2.97M, +810k weekly | Same kind, 30 countries |
| Shots - AI Photo & Video Maker(DeePix) | Android | 2025-07 | 2.55M, +350k weekly | Similar, 160k ratings |
| AI Face Aging(Willson Studio) | Android | 2026-02 | 7.13M, +570k weekly | Single feature (aging), matching Picsgen's Future Me |
| AI Art Magic / AI Photo Studio: Art Magic / AI Photo Editor: Art Studio, etc. | Android | 2026-05~09 | 100–4M, +25~500k per week | Nearly identical template wrappers, a large batch launched around the same time |
| AI Photo: Magic Editor pro (Nubivio Apps, Terafort sibling) | Android | 2026-08-28 | 108k | Same company, cloned under a different account |
| Google Gemini / Nano Banana, ChatGPT image generation | Big-company features | From ¥2025 | Hundreds of millions of users | Free and prompt-driven; Picsgen sells "no prompt writing needed" templates |
In our library, ranked by "photo + AI" install growth over the last 7 days, Picsgen's 208k doesn't even make the top 30 (#30 sits around 240k). This is a brutally crowded category, and Picsgen is just one of the mid-tier players.
7. Differentiation
Almost nothing. The feature list (image-to-image templates, image-to-video templates, restoration and enhancement, cartoonization, aging, muscle transformation, weekly trending effects) is nearly identical across the dozen-plus wrappers in the table above, and the backend is most likely a third-party image/video generation API. There are only two things worth noting: (1) template and keyword localization for India / South Asia (Couple in Rain, Love Moments, festival themes); (2) a studio with mature paid acquisition and a multi-account matrix running it. It grows on distribution execution (paid acquisition + ASO + reviews + account matrix), not on differentiation.
8. Takeaways
- Emerging markets (India, Nigeria, the Gulf, South Africa) can buy a million installs for AI template apps within weeks. But work out LTV first: judging by AppWill's books, a wrapper at this scale can run a negative monthly net profit.
- Custom store-page titles and localized templates per country, with descriptions stuffed with long-tail keywords.
- Copy one codebase across multiple accounts (copies from the same company, Nubivio Apps), then list the source code on AppWill to monetize. That's the full cash-out chain for studios like this: ads + in-app purchases + selling the source code.
- A cautionary tale: with no retention and no word of mouth, its Top Free rank dropped from 14 to 40 within a week.
9. Uncertain
- I couldn't get its ad creatives or ad spend (nothing in the Meta library; no Google or TikTok ads found). "Paid acquisition is the main engine" is an inference from the attribution SDKs, the country mix on the charts, and AppWill's negative profit.
- Whether the AppWill listing is actually this package's app is unverified (100% confirmed).
- Actual revenue: AppGoblin estimates $50K+/month, which is far off from AppWill's books (net -$1.3K to $2K per month). Net revenue may already be after acquisition costs; gross revenue is unknown.
- Whether the difference between the US five-star histogram and India's 4.2 stars comes entirely from fake reviews is unverified.