A UK "car rental company" runs an adult AI companion: Omi is just a front, and the real play is a reskinning playbook
What it is: An adult-oriented AI roleplay chat app built around a "late-night mood". Each message costs one coin, and there's also a subscription.
Why it took off: Not the product. The same backend gets rotated through reskins, paid ads run inside other companion apps to pull users in, and "give five stars, get coins" pushes the rating up to 4.6. It targets small-language charts in countries with low barriers, like Poland and Germany.
CHow you could follow it: Don't build the 61th reskinned companion app. If you have paid acquisition capability, build a companion with transparent pricing, a subscription that really is unlimited, and solid small-language translation. Make "no per-message charges" the selling point and go after the users these apps have been ripping off.
- Data (updated daily) · first launch 2026-09-03 · 11-07 installs 80,487
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)
Omi: Velvet Night hit 80k installs two months after launch and charted in Poland's and Germany's Entertainment Top Grossing. Check the developer account, the website and the privacy page, and it's one of several front brands a companion team ships in bulk: same backend, new name, new shell company, new domain, and once it reaches 70–80k installs, the next one ships. If you want to follow, the playbook is what's worth learning. The app itself has little to copy.
An ad seen inside another companion app
In 2026 Oct 5, a US user left a one-star review of Omi on Google Play: "I downloaded it because I saw the ad in Emochi. The ad said it's better than Emochi. But it's like 2000-era texting: every message costs real money in coins, and I got stopped after two messages." An earlier review from Sep 19: "The ad says no filter and no rules, but the community guidelines have rules."
Those two reviews pretty much explain where it came from: users are bought with ads, the ads run inside similar companion apps, and the pitch is "more freedom".
Now look at the developer on the store page: WT RENTAL CARS LIMITED. UK company registration shows it was incorporated in Leeds in Jun 2026 18, with car rental listed as its main business activity (Companies House 17284581). Less than three months after incorporation, it published an adult AI companion app, and this is the only app under its name.
What it is and how many installs it has
An AI character chat app built around a "late-night mood", with mostly anime-style characters and adult-oriented images, video and voice. The store description reads "AI character experiences for conversations after dark".
| Date | Total installs | Score | Rating |
|---|---|---|---|
| 10-29 | 63,969 | 2,262 | 4.64 |
| 10-31 | 73,213 | 2,379 | 4.60 |
| 11-03 | 78,040 | 2,568 | 4.41 |
| 11-06 | 80,487 | 2,794 | 4.68 |
Source: Google Play store data. Launched 2026-09-03, +8k installs in 16.5 days, about 2,000 per day on average.
11-06 ranks: Poland Entertainment Top Grossing #63, Top Free #78; Germany Entertainment Top Grossing #106; France, UK and Canada Entertainment Top Grossing between #140–160. Not on any overall chart. In-app purchases run from $2.99 to $799.99.
One number looks off: the rating count is about 3.5% of installs, where a typical app sits around 1%. The reason is below.
Where the money comes from: coins per message, and subscribers still pay per message
Each message costs coins. German users' reviews are the most direct: "I wrote 2 messages, then 'Sorry, you need to buy gems to continue'", "Even at over 44 euros a month, you can still only send a limited number of messages" (10-11). About 30 free coins are given out per day. A Polish user worked it out: "enough for 2–3 messages" (10-07). After subscribing, photos and videos still cost extra. One person said a photo is listed at 5–7 coins but actually deducts 15–30 (10-22).
The design is aggressive: a little free currency each day brings people back, and real conversations get cut off after a few messages, right when they're getting into it. Complaints are plentiful, but it still ranks in Germany's Entertainment Top Grossing top 110, so people are clearly paying. From the outside, actual revenue isn't visible. The rankings suggest it's not huge (inferred).
Where users come from: paid acquisition, plus "five stars for coins"
Paid acquisition. Apart from the two reviews above, a search on TikTok, X and in the media turns up no organic content about it at all. The only results are APK download sites. The store numbers don't rise smoothly either: 10-31 jumped +9,244,11-03 and then +4,827, with almost no movement in between, which looks like batches of paid installs (inferred; Play install counts themselves only update every few days).
Ratings for coins. The reviews make it plain: "Just started playing, giving five stars gets you coins" (10-04, in English); "I only wrote it for the points XD" (in Polish); "There are lots of ways to earn free coins." On top of that, plenty of bare five-star reviews (nice, cool, git, super). The rating count is more than three times that of similar apps, with the score holding around 4.6, which helps paid-acquisition conversion. This violates Google Play's review policy.
Targeting small-language markets. The main battlegrounds are Poland and Germany, not the US. Polish users praise that "the Polish is quite complete", though some say "the translation sometimes gets garbled". The Entertainment Top Grossing charts in both countries have low entry barriers, and paid acquisition is cheaper than in the US.
It's actually a shell brand
Use the developer site's look as a lead: www.<name>.top, the privacy page at /privacy-policy.html, and a Gmail contact address. That turns up about 30 AI companion apps, and they all look the same:
- Developer names have nothing to do with the business: car rental (WT RENTAL CARS), automotive (FIRST MILE AUTOMOTIVE), food (VIA DEL GUSTO), logistics, real estate, diamonds, plus a batch of individual accounts in Egypt, Bangladesh, India and Nigeria.
- The privacy policy is the same text with a different name: "Omi is an AI-powered interactive application designed around personalized conversations…" Auri, JulyGF and Echora open with nearly identical wording. One site's page source still carries traces of text copied from a Feishu doc, so the team is probably based in China (inference).
- The websites use one template: the homepage is just a few slices exported straight from a design file.
- Each domain gets its own server: we checked 18 domains and found 18 distinct IPs, all on AWS us-east. The point is that Google can't see they're one company and ban them all at once.
- The in-app purchase ranges are identical: Omi, Echora and Auri all run $2.99–$799.99.
They launch in rotation: Echora went live in Aug 20, pushed to 78k installs, then dropped off the charts; Omi picked up the baton in Sep 3; JulyGF launched in Sep 10, reached only 1,330 installs and never took off; Auri launched in Oct 15, went from 10-31 to 11-05, rising from 2,773 to 10,432, and is now taking the next turn.
Users have noticed too. One review says "exactly the same as sealove, just the accounts don't carry over", another says "the UI isn't very friendly, it's a step backward from the last app" (10-02).
Who it's competing with
Almost everyone. In the 11-06 German Entertainment Top Grossing top 200, there are about 60 companion and roleplay apps with AI, Chat or Character in the name, and most of them are much bigger:
| Product | Launched | Installs (Play, 11-06) | vs. Omi |
|---|---|---|---|
| Character AI | 2023-05 | 70.7M | Big tech, mostly free, strict content filtering |
| Talkie | 2023-06 | 60.8M | Chinese big tech going global |
| PolyBuzz | 2023-06 | 57.9M, 7 days +370k | German Entertainment Top Grossing #11, the top of the Chinese-team global pack |
| Emochi | 2024-01 | 29.7M, 7 days +660k | Omi's ads run inside it |
| Chai | 2021-03 | 24.4M | Veteran, also per-message / subscription |
| Afizzy | 2026-08 | 1.33M, 7 days +110k | Launched the same period as Omi, 15× bigger |
| Mira(MINDZEN AI) | 2026-09 | 320k, +84k in 7 days | Same company, multiple apps |
On the web there are adult companion sites such as JanitorAI, SpicyChat and Candy.ai, which are completely outside store review.
As for the product itself: one user says "it's not one of those 20k clones", while another says it's identical to sealove. Looking at it, the differences are only packaging: a "late night, velvet" tone, adult images and video, voice, and a choice of model tiers. Dozens of similar apps on the German chart already do all of this.
If it were you
Verdict: Nobody knows this app, yet the playbook it uses is one everyone is already running. Adult AI companions plus a shell-app portfolio plus paid user acquisition are crowded, and you have no edge in there. There's something to learn, but following it isn't worth it.
Who it suits: teams already running an overseas companion app or another subscription tool, with a backend they can reuse, a paid acquisition team, and spare developer accounts. For you, the things worth taking are "start with small-language markets like Poland and Germany where localization is weak and chart thresholds are low" and "give out a little currency daily to keep daily active users up."
Who it doesn't suit: indie devs without an ad budget or paid acquisition experience. The users here are all bought. Push one app to 80k installs and you need ad spend; you have no organic traffic to lean on and no community to fall back on.
If you must build: don't make the 61th generic companion app, and don't make another reskin. Try the gaps reviews keep pointing to: users hate "charged per message even though I subscribed", "5 coins charge 30", and "account banned after cancelling." A companion app in a smaller language with transparent pricing, a truly unlimited subscription and solid translation, selling "no per-message fees" as its ad hook and targeting users of big apps like Emochi and PolyBuzz, is the only angle that holds up. You still have to buy users, mainly through in-app ad networks and Meta.
When was the window: there isn't a clear one. This business doesn't run on a single event. Google could clear out reskinned apps and rating-farmed apps in bulk at any time, and if it does, it would open room for the people doing it properly.
What would show you're wrong: if the Auri portfolio keeps pushing a new app up every few weeks without mass removals, that would mean reskinning plus rating-farming works on Play over the long term. Then it's a steady business, not just a grey-zone arbitrage. Conversely, if Omi and Auri both get pulled within a few weeks, the route is being closed off.
Research notes: competitor table, acquisition evidence, full research report
- Competitors
Name Where Launched Scale How it differs from this one Character AI iOS 2023-05 Play 70.7M installs Big tech, mostly free, strict filtering; Omi is a small adult, pay-per-message app PolyBuzz Android 2023-06 Play 57.9M, 7-day +370k, German Entertainment Top Grossing #11 A single top Chinese global brand, two orders of magnitude bigger; Omi is a one-off reskin inside a portfolio Emochi(FlowGPT) Android 2024-01 Play 29.7M; 7-day +660k Has a creator community; Omi has none and relies purely on paid acquisition Talkie Android 2023-06 Play 60.8M Big-tech global brand Chai Android 2021-03 Play 24.4M Veteran, also per-message / subscription, far larger scale Afizzy - AI Character Chat Android 2026-08-06 Play 1.33M; 7-day +110k New app launched the same period, with about ten times Omi's paid acquisition scale Mira - AI Chat & Companion(MINDZEN AI) Android 2026-09-18 Play 320k, +84k in 7 days Same company, multiple apps, but advertising under its own brand JanitorAI / SpicyChat / Candy.ai Web 2023–2024 Not verified Adult companion on the web, not limited by store review - Where users come from
- Inferred Paid acquisition (inferred): no organic content at all (TikTok, X, Reddit and media searches come up empty); install counts jump in batches (10-31 +9,244, 11-03 +4,827); our Meta / TikTok / Google ad libraries don't index it, so we can't pull the creatives
- Confirmed Shell-portfolio reskin: the developer is a UK car-rental shell company registered to 2026-06; about 30 AI companion apps share www.<name>.top/privacy-policy.html, a Gmail address, the same design-slice homepage and the same AI-generated privacy policy text, each domain on its own AWS EC2; Omi, Echora and Auri all run in-app purchases at $2.99–$799.99
- Confirmed Incentivised ratings (coins for tasks): reviews include "wanted the reward", "robię to za punkty XD" (I'm doing it for the points) and "lots of different ways to earn free coins"; the ratings-to-installs ratio is about 3.5%, more than 3× a typical app
- Confirmed Localized push into small-country categories: the main battlegrounds are Poland (Entertainment Top Grossing #63 / Top Free #78) and Germany; Polish reviews praise the complete Polish localization, German reviews say the translation is sometimes off
Why Omi: Velvet Night (com.ominight.vevlet) could chart
Channel: Google Play · Category ENTERTAINMENT · Developer "WT RENTAL CARS LIMITED" · Store debut 2026-09-03 (earliest review 2026-09-25) · Research day 2026-11-06
Bottom line
It isn't an "independent hit". It's the skin the portfolio is currently pushing from what is most likely a Chinese overseas AI companion team. The same backend and monetization (coins deducted per message, plus subscriptions) gets a new name, a new developer account and a new .top domain every few weeks. Paid acquisition pushes it to 5–100k installs, then "complete tasks for free coins (including rating the app)" drives the rating up to 4.6+. It gained traction through distribution tactics (paid acquisition, portfolios, incentivised ratings, aimed at low-competition localized markets like Poland and Germany), not because the product is different.
1. The numbers (source: our googleplay database)
| Date | Total installs | Score | Rating |
|---|---|---|---|
| 2026-10-29 | 63,969 | 2,262 | 4.64 |
| 2026-10-31 | 73,213 | 2,379 | 4.60 |
| 2026-11-03 | 78,040 | 2,568 | 4.41 |
| 2026-11-06 | 80,487 | 2,794 | 4.68 |
- 8 days added 16,518 installs (about 2,000 a day), and 532 new ratings. The ratings-to-installs ratio is ≈ 3.5%, more than 3× the typical app (about 1%), explained below.
- Chart positions (2026-11-06): Poland Entertainment Top Grossing #63, Top Free #78; Germany Entertainment Top Grossing #106, Top Free #132; France #141, UK #145, Canada #158 (all category charts, none on an overall chart). 10-31 to 11-01 it also briefly appeared at #192–197 on the US Entertainment Top Grossing chart.
- In-app purchase range $2.99–$799.99; the store lists ad_supported = True.
- This is the only app under the developer account.
How to read revenue: the top 100 on the German Entertainment Top Grossing chart are almost all AI companions (PolyBuzz #11, Crushie #17, Emochi #19, Chai #24…), with Omi at #106. The Polish chart has a lower bar. That shows real paying users, but the volume is small. Exact amounts are unverified.
2. Competitors: this niche is already packed
On 11-06, the German Entertainment Top Grossing top 200 included about 60 AI companion and roleplay apps with AI, Chat or Character in the name. A few representatives:
| Product | Where | Launched | Scale (googleplay library, 11-06) | How it differs from Omi |
|---|---|---|---|---|
| Character AI | Android / iOS / web | 2023-05 | 70.7M installs | Big-tech scale, mostly free, strict content filtering; Omi is adult-oriented and pay-per-message |
| PolyBuzz(Cloud Whale) | Android / iOS | 2023-06 | 57.9M, 7 days +370k | Top Chinese global brand, German Entertainment Top Grossing #11; two orders of magnitude bigger |
| Talkie(SUBSUP / MiniMax) | Android / iOS | 2023-06 | 60.8M | Chinese big tech going global |
| Emochi(FlowGPT) | Android | 2024-01 | 29.7M, 7 days +660k | Some reviewers call themselves "emochi pro content creator", so the user bases overlap |
| Chai | Android / iOS | 2021-03 | 24.4M | Veteran, per-message / subscription |
| Afizzy(Guru AI Lab) | Android | 2026-08-06 | 1.33M, 7 days +110k | Launched the same period as Omi, 15× bigger |
| Mira(MINDZEN AI) | Android | 2026-09-18 | 320k, +84k in 7 days | Also a new app from the same period; MINDZEN runs multiple apps too (Fantasia, Rosytalk) |
| SeaChat / Sea Soul(SeaArt) | Android | 2026-09 / 2026-03 | 50K / 6.17M | SeaArt has its own image-generation capability |
| Whispy AI(NEXBASE) | Android | 2026-06 | 550k | UK Entertainment Top Grossing #13 |
| Web: JanitorAI, SpicyChat, Candy.ai, etc. | Web | 2023– | Not verified | Web-based adult companion chat, not subject to app store review |
Sibling packs from the same matrix (see next section), like Echora, Auri, JulyGF, are really its "competitors" too. It's just the same company competing with itself for traffic.
Bottom line: Tons of products do the same thing, and most are bigger. At the product level, Omi has nothing visibly unique.
3. Key finding: this is a skin matrix
Searching the googleplay library for AI companion apps whose developer site is https://www.<name>.top/, privacy page at /privacy-policy.html, and contact email is Gmail, turns up about 30 apps. They share:
- Developer names unrelated to the business: WT RENTAL CARS LIMITED (car rental), FIRST MILE AUTOMOTIVE LTD (auto), VIA DEL GUSTO (food), Praise shipping & logistics (shipping), BOLTEC PROPERTIES LIMITED (real estate), Veer diamond Limited (diamonds), Auction Point, dera Nigeria limited, Ecoplastile... plus a batch of individual developer names from Egypt, Bangladesh, India, and Nigeria.
- WT RENTAL CARS LIMITED is a shell company in Leeds, UK, registered 2026-06-18, SIC 77110 (car rental), Companies House number 17284581 (TinyTax). It listed this app less than 3 months after registration, which looks like a bought or rented developer account.
- A separate AWS US-East EC2 per domain (omivelvetnight.top 3.89.41.221, echorale.top 3.217.72.151, whispergf.top 34.206.32.156, camilax.top 98.94.162.91... 18 domains on 18 different IPs, all nginx). This is deliberate isolation to keep the apps from being linked and taken down together by Google.
- Same homepage template: the homepages of omivelvetnight.top, echorale.top, and camilax.top are all just a few
FigmaDDSSlicePNG<hash>.pngslices (static pages exported straight from design files). - Same privacy policy template, AI-generated:
- Omi:"Omi is an AI-powered interactive application designed around personalized conversations, atmospheric digital experiences, and evolving virtual interactions."
- Auri (listed 2026-10-15, developer VIA DEL GUSTO): "Auri is an AI-powered interactive application designed around personalized experiences, adaptive…"
- JulyGF: Velvet Days(2026-09-10,VixonD):"JulyGF is an AI-powered interactive application designed around ongoing conversations, personalized virtual interactions…"
- Echora(2026-08-20,FIRST MILE AUTOMOTIVE LTD):"Echora is an AI-powered interactive experience built around evolving conversations…"
- All four sites share the same
/user-policy.html. The whispergf.top privacy page HTML still containsace-line old-record-id-…, a trace of content pasted from Feishu docs, suggesting the team uses Feishu and is probably based in China.
- Identical in-app purchase ranges: Omi, Echora, and Auri all use $2.99–$799.99 (googleplay library app_snap).
- Rotation cadence: Echora (listed 08-20, at 10-29 had 78k installs and 3,193 ratings, then dropped off all charts) → Omi (listed 09-03, traffic started late September) → JulyGF (09-10, only 1,330 installs, never took off) → Auri (listed 10-15, went from 10-31 to 11-05 and rose from 2,773 to 10,432). Each pack gets pushed to around 80k, then swapped for the next.
- One user review says outright: "same as sealove. (seelove?) I don't think they are transferrable tho." (Same as sealove, just the accounts don't transfer.) Users themselves can see this is the same product in a new skin.
4. Where users come from
| Channel | Evidence | Confirmed? |
|---|---|---|
| Paid acquisition (Meta / Google / TikTok, etc.) | No organic content at all (a TikTok search for "Omi Velvet Night" returns 0 related videos; x-radar shows 60 posts in the last 0 days; Exa / Jina only surface APK download sites). Installs jump 9,244 and 4,827 within three to four days, which matches ad-batch timing. The matrix playbook is built on paid acquisition. Our Meta / TikTok / Google ad libraries don't include it (it's not on our tracking list), so we can't pull specific creatives. | Inferred, not directly confirmed |
| Matrix reskins + localization for small markets | See the previous section. Main battlefields are Poland (Entertainment Top Grossing #63 / Top Free #78) and Germany. Polish reviews say "bardzo rozbudowana polski język" (the Polish language support is very thorough), though some say translations sometimes break; German reviews say translations aren't always good. | Confirmed |
| Incentivized ratings (rewards for tasks in coins) | Reviews: "wanted the reward, ai couldn't even get its own name right"; Polish "robię to za punkty XD" (I'm only doing this for points); "lots of different ways to earn free coins"; ratings count / installs ≈ 3.5%; lots of one-word five-star reviews ("nice" "cool" "git" "super"). | Confirmed |
| Store search | Titles stuffed with Velvet Night / AI characters / Late-night AI conversations, and screenshots lean adult. | Weak, unverified |
| Influencers / media / communities | None found | None |
5. Why users stay and pay (from the latest 80 reviews each in the en / pl / de stores, pulled with google-play-scraper)
- They want adult content: "there is a lot of nude stuff", "this app is visually very adult", "nice naughty characters and pictures". Photos and videos require separate paid unlocks, and some paying users complain they had to pay extra after buying a subscription.
- Chat quality and voice are decent: "responses are very detailed, even on the cheapest bot setting", "the voices sound amazing", "memory is quite good". Users can pick different models, and cheaper and pricier tiers cost different amounts of coins.
- Monetization is aggressive: coins are deducted per message. "5 Nachrichten keine coins mehr" (out of coins after 5 messages); "Über 44 Euro im Monat und man kann dennoch nur begrenzt Nachrichten schreiben" (over €44 a month and still limited messages). Others report actual deductions exceed the listed amounts (listed 5–7 coins, actually charged 15–30), and accounts getting banned after canceling subscriptions. About 30 coins are given out daily, roughly enough for 2–3 messages, which builds a habit of coming back every day.
- The content skews anime, and many people want realistic characters.
6. What's the differentiation, and is it why it broke out?
- No visible product differentiation. One user says it's "not one of those 20k clones," yet another says it's identical to sealove. The privacy page, homepage, and in-app purchase structure all match the sibling packs.
- If you have to name a difference, it's only packaging: a mood-driven "late night / velvet" positioning, adult images and video, voice, selectable models, and localization into Polish and German. In the German chart, these are standard features shared by dozens of similar apps.
- It works because of the playbook, not differentiation: a reusable backend and monetization template, one pack pushed to volume before swapping to the next skin, changing developer accounts and domains to spread risk, paid acquisition, coin rewards for reviews, and targeting Poland and Germany for their lower competition and cheaper ads.
7. How many people know it
- No media coverage, nobody on X talks about it, no TikTok content, no TrustMRR listing or founder revenue posts. The only results are APK download sites (Softonic, FileHippo, CNET, APKPure, AppBrain) and auto-generated Apptopia pages.
- But the business itself (overseas AI companion apps + matrix of skins + paid acquisition) is already crowded: the
.toptemplate alone covers about 30 packs, and about 60 similar apps sit on the German Entertainment Top Grossing chart. This single product went unnoticed, but the playbook is a public red ocean.
8. What others can borrow
- One backend, many front-end skins: each skin gets a new name, theme, characters, and developer account, rotated in turn, so taking down any single pack doesn't hurt the operation.
- Go after Poland and Germany first, in categories with low localization and low Top Grossing bars, instead of head-to-head in the US.
- Per-message coin deductions plus subscriptions, with a small daily coin giveaway and a task wall (including rating for coins) to drive daily actives and inflate ratings.
- Risks: skin matrices and incentivized ratings both violate Google Play policy (Repetitive Content, Ratings & Reviews), so mass account bans are possible. Adult content carries policy risk too. Users complaining about opaque charges and accounts banned after canceling mean refund and bad-review risk is real.
9. Still unclear
- Which platforms it buys on and how much, our ad libraries don't cover, so we can't confirm.
- Who is behind it. Feishu traces and the design-slice template point toward a China-based team, but there's no hard proof.
- Revenue or retention. A Top Grossing category rank only shows that people pay, not how much.
- We couldn't identify which app sealove / seelove is (the name isn't in the googleplay library, so it may be delisted or renamed).
- The store's rating distribution swings wildly day to day (1 stars going from 16 to 153 and back to 78). It could be Google cleaning up reviews, or someone brigading with bad reviews. Cause unknown.
Sources
- Our googleplay library (agent-im server DuckDB): app, app_snap, app_day, chart_rank, app_profile, 2026-10-29 to 11-06
- Google Play page: play.google.com/store/apps/details?id=com.ominight.vevlet
- Reviews: google-play-scraper pulled the latest 80 each from en-us / pl-pl / de-de (2026-11-06)
- Company info: tinytax.co.uk/company/wt-rental-cars-17284581 (Companies House 17284581)
- Privacy pages: www.omivelvetnight.top/privacy-policy.html, www.echorale.top/privacy-policy.html, www.cwauri.top/privacy-policy.html, www.julygf.top/privacy-policy.html, www.whispergf.top/privacy-policy.html
- DNS resolution (getent, 2026-11-06)
- TikHub TikTok search, x-radar, Exa, Jina search (2026-11-06); our metaads / tiktokads / googleads libraries (no results)