An AI video shell from Da Nang, Vietnam, gained 8,731 installs in one day: what took off isn't the product, it's the distribution machine behind it

On Nov 2, an app called "AI Video - AI Photo Generator" on Google Play gained 8,731 installs in one day, and then gained none for the next two days. It had been listed for less than a month, with no ratings and no reviews, yet it was already appearing on the free Photography charts in 7 countries including Italy, Nigeria and South Africa. It looks like it's gaining traction, but it's actually a Vietnamese tool-app publisher's reskinned package. Below: what it is, where the installs come from, who is running this business, and whether you should follow it.

What it is

An Android AI video app published by TOMODEV JSC, a company based in Da Nang, Vietnam, listed on Oct 9. The store page lists four features: text-to-video, photo-to-animation, text-to-image, and multi-photo slideshows. The store pitch is one line: "Turn your ideas into moving stories", followed by "No advanced editing skills needed."

It makes money two ways: the app shows ads, and it also sells $7.99–$31.99 in-app purchases (Google Play store page, 11-06). This "ads as the floor, subscriptions or credits as the bulk" model is common among tool apps on Play.

Nothing about the features looks different from competitors. The last one, the "photo slideshow," has nothing to do with AI; it looks lifted from an ordinary video-editing template.

Did it really take off?

It has some traction, but it's far from a hit.

Source: Google Play install counts, recorded daily, 10-29 to 11-06.

It gained 16k in a week. For the AI video category that's small: Nexus - AI Video Generator, a peer in the same category, gained 1.85M in the same week, more than a hundred times as much. It has only charted on the Photography free chart, its best showing being #88 in Italy (11-06), and it has never appeared on the overall or top-grossing charts.

The curve's shape says more than the numbers. It rises for two days, stalls for two, then rises again, like steps rather than word of mouth spreading gradually. It looks more like installs released in batches from ad spend (inferred).

Where the traffic comes from

1. It belongs to a matrix of dozens of apps. The developer website this app lists on the store is not its own site; it is lutech.vn/app-ads.txt, the ads configuration file the Vietnamese publisher Lutech Digital uses to serve ads across its apps. The number of Google Play apps listing lutech.vn as their developer website is 39, spread across at least 8 developer accounts including TOMODEV JSC, ZAppsolution, SHINK COMPANY, JoyNest Studio and BG.Studio, with most contact emails at [email protected]. The categories are all proven tool-hit themes on Play: emoji battery (12.7M installs), font keyboards (24.4M), IPTV, plant identification, pregnancy, short drama, football scores, and AI video. Lutech's website says it has been "2024 years among the world's top 200 publishers by downloads," has "100 game and app studios," and presented a case study at the TikTok Apps Summit SEA 2025 (lutech.vn, 11-06).

2. The same theme has been tried at least three times. The same account published GlimAI - AI Video Generator in June 2025, now at 410k installs; ZAppsolution published an almost identically named "AI Photo - AI Video Generator" in March 2026, now at 1M installs, with the contact email also [email protected]. This package is the latest reskin-and-retest of the first two. The earlier ones reached hundreds of thousands and millions of installs, which suggests this theme pays back for them.

3. They're likely buying installs, but no ads for this package were found. In the TikTok ad transparency library, the Vietnam-registered "LUTECH SOFTWARE TECHNOLOGY COMPANY LIMITED" has been running ads in Europe since September 2025, and 11-06 has 192 active ads, promoting emoji battery, English speaking and other packages. No ads for this AI video package matched, but only the European TikTok ad library was checked, so absence isn't proof it wasn't run. The countries where it charted include Nigeria, South Africa and the Philippines, as well as Italy, Poland, France and Spain: test cheap emerging markets for volume first, then try a few European countries. This is the usual route these publishers take when testing new packages with Google UAC and TikTok (inferred, not verified).

4. Titles and package name stuffed with keywords. The title jams 'AI Video' and 'AI Photo Generator' together, and the package name aiphoto.videogen.aigenerator.enhance is all search terms too. It's chasing long-tail traffic from store search (inferred, no keyword ranking data).

What's missing: Searching for it online turns up only the Google Play page and a few APK mirror sites. No press coverage, no Reddit or X discussion, no user reviews.

Who it's competing with

AI video generation is one of the most crowded categories on Google Play. 243 apps with titles containing ai video, video generator, photo to video, or image to video; 62 of them launched after August 2026, and 84 are growing faster than it week over week (Google Play, 11-06).

In this table it's a tail-end newcomer. Above it sit players with their own models, faster-growing rivals, and a pile of free big-tech products overhead.

What to learn from it

What's actually worth studying is Lutech's playbook, not this app:

• Spread risk across accounts. Dozens of packages spread across eight or nine developer accounts, so one account getting banned doesn't take down the rest.
• Re-skin the same concept over and over. The concept is ready-made, the shell is ready-made. Change the name and icon and you have a new package. If it runs, spend more; if not, drop it.
• One app-ads.txt routes all ad monetization. Ad revenue backs the bets, letting them buy installs in places like Nigeria or the Philippines where a single user is worth little (inferred).
• Packing package names and titles with keywords.

For a publisher with an in-house paid acquisition team and an ad monetization hub, these are an assembly line. For a small team, each piece has to be built from scratch.

If it were you

The verdict: Don't go head-on. This package broke out because of the group's paid acquisition, monetization, and multi-account matrix, not the product itself, and those are exactly what followers can't access. In this AI video shell category, the leader's weekly new installs are a hundred times this package's total.

Who it suits: Teams already running a matrix of Play utility apps that have paid acquisition and ad monetization capability. You already know how this works. AI video is just one row on the opportunity list; test it the same way.

Who it doesn't suit: Solo indie devs who can code but have never run UAC or TikTok ads; anyone hoping to start from organic store traffic or SEO; anyone trying to win the category with a 'better AI video app.' Even a great product will struggle to get seen here.

How to do it (if you must touch it): Don't build the #244 generic AI video wrapper. Go narrow: pick one specific use case (animating old family photos of someone who has passed away, wedding invitation videos, pet birthday videos), build one template and make it the best at that, and capture search with scene-specific keywords. Or skip building an app altogether and copy the playbook instead: multiple accounts, reskinned variants, test in low-cost countries, and apply it to a tool niche that isn't crowded yet.

How to get users: For a narrow scenario, the creative is the finished video. A vertical video of 'old photo comes to life' will do more than any feature list. Start with a small budget in one or two cheap countries and test 20–30 creatives; only scale spend once install costs are manageable (inferred).

How much time you have: No rush, and this window belongs to this publisher, not you. If you go narrow, you have roughly a few months (estimate) before big players build the same templates into Gemini or the free tier of CapCut.

Signs this read is wrong: ① This package gained 500k installs in a month and starts getting real user reviews, which would mean it has organic traffic we haven't seen; ② it reaches a country's Top Grossing chart, which would mean people really buy in-app purchases and the business earns more than it looks; ③ an indie dev without paid acquisition makes verifiable, public revenue from an AI video shell.

Data and sources

• Installs, charts, IAP prices: Google Play store pages and daily install counts, 10-29 to 11-06.
• Number of peers and rival scale: Google Play apps with AI video-related terms in their titles, 11-06.
• Publishing setup: developer websites and contact emails for each app on Google Play; lutech.vn, tomodev.vn (11-06); the TOMODEV JSC developer page on AppBrain; the ZAppsolution publisher page on AppMagic (attributed to TOMODEV JSC). Whether TOMODEV and Lutech are a holding structure, a co-publishing deal, or the same people is unverified. We can only confirm they sit in the same publishing setup.
• Ads: TikTok Ad Transparency Center (Europe only), 11-06. No match for this package in the Meta or Google ad libraries.
• Where you see 'inferred' or 'estimate', that's an inference or an estimate.