5 AI wrappers: the flip week
This week’s teardown covers Ploxto, Roman AI, Elofoot, VidTL, and ShortClip — five AI wrapper SaaS products split between verified-revenue marketplace listings and smaller build-in-public bets.
[data] This week's screen produced five AI wrapper SaaS products with public traction: Ploxto, Roman AI, Elofoot, VidTL, and ShortClip. [editor's view] The cleanest read is a split market. The TrustMRR products have stronger verified revenue but weaker founder narrative. The build-in-public products have tiny MRR but clearer lessons for anyone trying to clone the workflow.
[data] Ploxto is already at $2,577 MRR from 310 active subscriptions, with a TikTok-led acquisition story that reportedly cleared 60 million views. 1 [data] Roman AI is at $5,425 MRR with only 6 active subscriptions, which makes its pricing and enterprise angle more interesting than its raw user count. 2 [data] Elofoot is the biggest top-line product in the batch at about $9,985 MRR, but it is also tied to a time-sensitive 2026 World Cup window. 3
Three takeaways before the teardown:
- [editor's view] Distribution beats model choice. None of these products wins because the AI layer is hard to copy. The hard part is TikTok reach, Slack workflow trust, sports-betting timing, browser-video performance, or local-market positioning.
- [editor's view] The easiest clones are not the highest-revenue ones. VidTL and ShortClip have cleaner first steps for an indie dev, even though their current MRR is only $58 and $46 respectively. 45
- [editor's view] For-sale listings are useful but biased. TrustMRR gives verified revenue snapshots, but this week's three TrustMRR picks are all businesses being sold, so the sample tilts toward products founders are willing to hand off.
Replication scoreboard
More stars mean harder to copy, not better.
| Product | Technical lift | Information edge | Capital needed | Legal risk | Clone read |
|---|---|---|---|---|---|
| Ploxto | ★★☆☆☆ | ★☆☆☆☆ | ★★☆☆☆ | ★★☆☆☆ | Copyable product, hard channel |
| Roman AI | ★★★★☆ | ★★☆☆☆ | ★★★☆☆ | ★★☆☆☆ | Integration-heavy, trust-heavy |
| Elofoot | ★★★☆☆ | ★★★☆☆ | ★★★☆☆ | ★★★☆☆ | Good timing, crowded niche |
| VidTL | ★★★★☆ | ★☆☆☆☆ | ★★☆☆☆ | ★★☆☆☆ | Hard browser engineering, clear wedge |
| ShortClip | ★★★☆☆ | ★★☆☆☆ | ★★☆☆☆ | ★★☆☆☆ | Best localization lesson |
1. Ploxto: AI photo editing with TikTok as the engine
[data] Ploxto is an iOS AI photo editor that wraps the Nanobana AI image editing API, so users can describe a photo change in natural language instead of manually editing the image. 1 [data] The app is published by ADVANCORIUM S.L., requires iOS 17.0 or later, and shows a 4.6 App Store rating from 43 ratings. 6
[data] The traction is real enough to study: $3,713 in last-30-day revenue, $2,577 MRR, 310 active subscriptions, and RevenueCat verification. 1 [data] TrustMRR lists the product for sale at $60,000, with 14 offers and 5,014 buyer views. 1 [data] App Store in-app purchases range from $6.99 to $39.99. 6
[data] The acquisition channel is the part worth stealing, if you can. Ploxto's listing says TikTok organic content produced more than 60 million total views. 1 [data] A Viral TikTok Tracker post reported one Ploxto video at 1,328,949 views. 7
[editor's view] The job-to-be-done is narrow and good: turn a visual edit request into a finished image without opening a pro editor. The clone risk is that the product is easy to describe and probably easy to ship, while the repeatable TikTok machine is not.
If you wanted to copy this: start with one TikTok-native transformation, such as outfit swap, room restyle, or dating-profile photo cleanup. Ship the iOS flow after you prove one video format can get saves and comments. The first likely failure mode is building a broad editor before you know which visual trick spreads.
2. Roman AI: the Slack coworker with enterprise-shaped revenue
[data] Roman AI is an AI coworker that lives inside Slack, connects to more than 3,000 tools, and uses persistent memory to execute tasks across a company's stack. 2 [data] Its website says, "Roman lives in Slack, connects to your entire stack, and actually finishes tasks, instead of giving you suggestions to copy-paste." 8
[data] Roman AI reports $8,828 in last-30-day revenue, $5,425 MRR, 6 active subscriptions, and Stripe verification. 2 [data] TrustMRR lists the business for sale at $90,000, discounted from $150,000, with 11 offers and 3,451 buyer views. 2 [data] The public pricing has a free tier, a $50 per month Pro plan, and Enterprise pricing starting at $10,000 per month. 2
[data] The tool connects to services such as Apollo, Crunchbase, LinkedIn, ZeroBounce, Instantly, PostHog, Notion, Google Drive, Gmail, Intercom, HubSpot, and GitHub. 2 [data] The website also promises "5-minute setup. No engineering required." 8
[editor's view] This is the least weekend-cloneable product in the batch. The wrapper is not just a prompt box. The hard work is permissioning, integrations, error recovery, and convincing a team to let a Slack bot touch real business systems. The 6-subscription count also means one or two enterprise-ish customers may dominate revenue.
If you wanted to copy this: do not start with "3,000 integrations." Start with one painful Slack workflow for one buyer, such as weekly sales-account prep from HubSpot and Gmail, or support escalation summaries from Intercom and Linear. The first likely failure mode is making a demo agent that looks smart but cannot safely complete a real task.
3. Elofoot: AI sports prediction timed to the World Cup
[data] Elofoot is an AI-powered football prediction platform that forecasts match outcomes, exact scores, and scorers across 46 competitions. 3 [data] The product uses xG, form, and injury analysis, and it was founded in May 2026 by solo founder Maxence Bombeeck. 3
[data] Elofoot reports $6,995 in last-30-day revenue, about $9,985 MRR, 411 active subscriptions, 20,000-plus active fans, and Stripe verification. 3 [data] TrustMRR lists Elofoot for sale at $85,000, with 4 offers and 808 buyer views. 3 [data] The stack includes Next.js, TypeScript, SwiftUI, Node.js, PostgreSQL, Docker, Railway, Stripe, and RevenueCat. 3
[data] Bombeeck says the team is selling because it does not have time to scale the product, and he frames the 2026 World Cup as the window to capitalize on a large sports-betting audience. 3 [editor's view] That timing cuts both ways. The World Cup can lower acquisition friction, but it can also make revenue spiky and push the product into a crowded prediction niche.
[editor's view] The cloneable part is not "AI predicts football." That pitch is everywhere. The better wedge is a local fan segment, a league-specific content loop, or a betting-adjacent use case that stays clear about what the product is and is not promising.
If you wanted to copy this: choose one league or tournament, publish free prediction content daily, and use email or Discord to test whether fans return before asking them to pay. The first likely failure mode is weak prediction credibility. If users cannot see why your output is better than a generic ChatGPT answer, churn will be fast.
4. VidTL: browser-native AI video editing
[data] VidTL is an AI video editor that runs in the browser, with local FFmpeg processing instead of server-side video upload. 9 [data] The product lists AI transcription, conversational editing, automatic subtitles, text-to-speech, voice cloning, silence removal, and multi-camera synchronization. 9
[data] The founder started the public build-in-public push on June 3, 2026, writing: "i love building. i hate marketing. but my product will not market itself." 10 [data] By Day 15, the founder reported $58 MRR and 45 users. 4 [data] VidTL pricing is Hobby free, Pro at $29 per month, and Studio at $99 per month. 9
[editor's view] The technical lift is higher than a standard CRUD wrapper. Browser video processing means WebAssembly performance, file handling, export reliability, and UX work that users will notice immediately. The upside is clear: if the editor truly avoids uploads, privacy and speed become part of the pitch.
[editor's view] The acquisition path is also honest. The founder is using public progress updates because the product has not yet found a scalable channel. That makes the current MRR less impressive but the learning loop more visible.
If you wanted to copy this: build one browser-only workflow for one creator pain, such as removing silence from podcast clips or converting long screen recordings into captioned shorts. The first likely failure mode is trying to compete with full video editors before one narrow export flow feels reliable.
5. ShortClip: localized faceless video for the CIS market
[data] ShortClip is an AI faceless-video generator for Russian-speaking and CIS creators: users enter a topic, and the product generates a script, voiceover, b-roll, animated captions, and a short video output. 11 [data] The tool uses ElevenLabs for 12 premium voices, OpenAI TTS for 7 free voices, AI script generation, automatic translation into Russian, and YouTube-to-Shorts clipping. 11
[data] Founder Ilya Pashayan launched ShortClip for the CIS market on June 17, 2026, and wrote, "This isn't just another AI wrapper." 12 [data] He described it as "a fully localized, end-to-end video creation platform designed for local creators." 12 [data] On June 24, he reported $46 MRR, 3 paying subscribers, and more than 100 registered users. 5
[data] ShortClip's public pricing is in rubles: Lite at 549₽ per month, Starter at 1090₽ per month, Pro at 2890₽ per month, and Business at 5790₽ per month. 11 [data] The product uses Laravel, Remotion, custom ASS subtitle rendering, and high-speed media processing. 11
[editor's view] This is the best localization lesson in the batch. The underlying workflow is familiar: AI script, voice, stock-style footage, captions, export. The market choice makes it more defensible than another English-language faceless-video app.
If you wanted to copy this: pick a non-English creator market where short-form monetization advice, local voice quality, and payment methods are underserved. The first likely failure mode is shallow localization. Translating the UI is not enough if voices, captions, examples, prices, and onboarding still feel imported.
What this week says about cloneability
[editor's view] The highest MRR product, Elofoot, is not automatically the best indie clone. It depends on timing, trust, and prediction quality. Roman AI has attractive revenue per customer, but a serious clone needs integration depth and a clear buyer. Ploxto looks technically simple, but its TikTok machine is the real asset. VidTL and ShortClip are smaller, yet they expose clearer build paths: narrow the workflow, own a channel, and prove users will pay before broadening the editor.
[editor's view] If you are choosing one idea from this batch, do not copy the surface category. Copy the constraint. Ploxto's constraint is viral visual demos. Roman AI's constraint is completing one business task inside Slack. Elofoot's constraint is a time-boxed sports audience. VidTL's constraint is browser-native processing. ShortClip's constraint is local-market fit.
What to do tomorrow morning
Pick one product above and write a 20-customer validation list before you write code. For each prospect, write the exact current workaround, the channel where you can reach them, and the smallest paid promise you could deliver in 7 days. If you cannot name the channel, do not build the wrapper yet.
Cover image: image from Ploxto TrustMRR listing
References
- 1Ploxto - $3,713 last 30 days
trustmrr.com
- 2Roman AI - $8,828 last 30 days
trustmrr.com
- 3Elofoot - $6,995 last 30 days
trustmrr.com
- 4
- 5
- 6Ploxto: AI Photo Editor App - App Store
apps.apple.com
- 7
- 8Roman AI official website
getroman.ai
- 9Home | VidTL
vidtl.com
- 10
- 11ShortClip official website
shortclip.ru
- 12
This story was produced automatically by a channel. One sentence is all it takes for Neodrop to keep producing for you.