Three AI wrappers at $51–$6,276 MRR: TikTok edits, mascots, and prompt reverse-engineering (2026)

[data] Three products qualify this week: Bluff AI, Mascofast, and PromptReverse. TrustMRR records founding dates of March 24, July 12, and January 8, 2026, respectively, so each falls inside the past-12-month window. Those dates are public founding dates, not guaranteed app-store launch dates. All three list teams of one or two and show public revenue or active subscriptions. 1 2 3

TL;DR

[data] Bluff AI is the revenue outlier: TrustMRR shows $6,276 MRR and 201 active subscriptions, while its founder note claims roughly $7,000 MRR, 12,754 users, and zero ad spend. Its wrapper turns one uploaded photo into a template-driven edit, but the real asset is an automated TikTok content loop. Mascofast is the cleanest small build: describe a mascot, generate variants, and export animations. It shows $141 MRR, 16 active subscriptions, and plans from $9 to $89 a month. PromptReverse is the broadest system: reverse-engineer image or video prompts, then run them across multiple models. It shows $51 MRR and two active subscriptions alongside 10,500 monthly visitors, mostly organic Google traffic. 1 2 3 4 5 6

[editor's view] The practical split is simple. Bluff AI has the strongest demand signal and the highest trust and policy exposure. Mascofast is the easiest manual pilot because the buyer's output is visible and easy to judge. PromptReverse has the most defensible-looking workflow, but also the most expensive product surface to reproduce. None of the three proves that the model call itself is the moat.

Three takeaways

[data] Bluff AI's public acquisition evidence points to organic TikTok, with a founder-described automated agent producing and posting creative. 1
[data] Mascofast names X/Twitter and SEO as marketing channels, and its official page says it launched on Peerlist. 2 5
[data] PromptReverse reports 120+ live tools, 93 guided workflows, and 10,500 monthly visitors that are predominantly organic Google traffic. 3

Replication scoreboard

[editor's view] More stars mean harder to replicate. For legal risk, more stars mean more exposure, not better legal hygiene. These are editorial estimates from the public product surfaces and revenue snapshots. TrustMRR pages were updated on August 17, 2026. 1 2 3

Product	Technical lift	Information edge	Capital needed	Legal risk	Clone read
Bluff AI	★★★☆☆	★★★☆☆	★★☆☆☆	★★★★☆	[editor's view] Easy demo; trust, content policy, and distribution do the work
Mascofast	★★★☆☆	★★☆☆☆	★★☆☆☆	★★★☆☆	[editor's view] Best manual pilot; consistency and commercial rights matter
PromptReverse	★★★★☆	★★★☆☆	★★★☆☆	★★★★☆	[editor's view] Strong workflow; multi-model operations are the bill

Bluff AI: the template is simple, the TikTok machine is not

[data] Positioning: Bluff AI turns a selfie or other uploaded photo into a polished edit or prank in one tap. The official site describes a three-step flow: choose a template, upload a photo, and save or share the result. Its templates include outfit swaps, hair changes, and background edits. The job-to-be-done is concrete: help a Gen Z user make a postable image quickly without learning an editor or writing a prompt. 4

[data] Traction and team: TrustMRR shows $6,276 MRR, 201 active subscriptions, $13,826 in all-time revenue, and Stripe verification. It lists two founders, Jordan and Aron Atli Gunnarsson, and a March 24, 2026 founding date. The same page carries Jordan's note claiming $7,000 MRR in three months, 12,754 users, and $0 spent on ads. The verified snapshot and the founder's rounded claim are different measurements; keep both separate. 1

[data] Acquisition, pricing, and stack: TrustMRR names organic TikTok as the growth engine. The founder says an automated agent writes the script, generates the character, renders the video, and posts it, with seven countries localized and 69 live templates. The paid funnel starts with a $0.99 three-day trial and then $6.99 per week for 450 credits, plus consumable credits. The public stack field only names Stripe, while the privacy policy identifies kie.ai's Red Panda AI for generation, Supabase for storage, and Railway for hosting. 1 7

[editor's view] The API call is the easy part. A clone can start with one template and one image provider. The harder loop is choosing a format people recognize on TikTok, producing enough variations to keep publishing, and measuring which result converts a free user into a weekly payer. The founder's automation claim is a distribution asset, not a technical moat that a new entrant gets for free.

[editor's view] Information edge: Three stars. Template performance, localized copy, and a library of formats can compound, but none is exclusive by default. Legal risk: Four stars. Bluff's privacy policy says uploaded photos are stored in Supabase and sent to kie.ai, and its terms prohibit non-consensual intimate imagery, deepfakes, and uploading real people's images without consent. A clone needs consent, retention, moderation, and age controls before it needs another template. 7 8

If you wanted to copy this

[editor's view] First concrete step: pick one harmless, shareable template and manually sell 20 generations to a narrow audience before building automated posting. First likely failure mode: the result looks plausible in a demo but users refuse to upload faces, or the content loop cannot produce a repeatable stream of posts.

Mascofast: selling a mascot workflow, not a prompt box

[data] Positioning: Mascofast turns a text description into a mascot reference image, style variants, poses, and animated video or GIF exports. Its official workflow is: describe a character, generate it, then animate it. The job-to-be-done is to help an app or game team ship a recognizable character and fresh variations without hiring an illustrator and animator for every iteration. 5 9

[data] Traction and team: TrustMRR shows $141 MRR, 16 active subscriptions, $222 in all-time verified revenue, and Polar verification. It lists Dikshit Jain as the sole founder, about 350 users, and a July 12, 2026 founding date. The public page also lists 85% last-30-day profit margin and 1,569 visitors, but those are marketplace fields rather than a customer cohort report. 2

[data] Acquisition, pricing, and stack: TrustMRR lists X/Twitter, SEO, and a Peerlist feature as the public acquisition surfaces. The official page says Mascofast is live on Peerlist and sells three monthly plans: Starter at $9 for 72 credits, Growth at $29 for 288 credits, and Studio at $89 for 864 credits. It says unused credits roll over. TrustMRR lists Next.js, Tailwind CSS, TypeScript, Supabase, and Polar; the model provider is not named. 2 5

[editor's view] Mascofast has the clearest binary test of the three: can a customer get a mascot that stays recognizable across a variant and an animation? The product can look like a thin wrapper if every generation is a new character. The recurring value comes from preserving a brand's identity while producing enough poses and exports to support a real release cadence.

[editor's view] Information edge: Two stars. A customer's approved mascot library is useful context, but the public product does not show an exclusive dataset or distribution relationship. Legal risk: Three stars. Mascofast's terms give users commercial rights while saying outputs may resemble existing works and require user review. Its privacy policy says prompts and generated content remain stored while the account exists, and deletion requests are handled manually through X. Those terms are a warning to build rights review and deletion flows early. 9 10

If you wanted to copy this

[editor's view] First concrete step: choose one buyer type, such as mobile game teams, and manually deliver one mascot plus six approved poses and one short animation for five paid trials. First likely failure mode: each new generation drifts away from the approved character, so the buyer still needs a human artist to make the asset usable.

PromptReverse: the moat-shaped part is the library and routing

[data] Positioning: PromptReverse analyzes an image or video, produces a reusable prompt, and lets the user remix and generate in the same workspace. The official site lists image-to-prompt, video-to-prompt, image editing, image generation, and video generation. The job-to-be-done is to turn a reference a creator likes into a repeatable production recipe, then run that recipe across models without opening several accounts. 6

[data] Traction and team: TrustMRR shows $51 MRR, two active subscriptions, $286 in all-time verified revenue, and Stripe verification. It lists Mani as a one-person founder, a January 8, 2026 founding date, and 10,500 monthly visitors. The listing says those visitors are predominantly organic Google traffic. 3

[data] Acquisition, pricing, and stack: The pricing is pay-as-you-go: a $1 trial pack, then $5, $15, $39, and $79 credit packs, with no subscription and credits that do not expire. TrustMRR lists React, Tailwind CSS, TypeScript, Node.js, Express, Cloudflare, Google Cloud, PostgreSQL, OpenAI, and Stripe. The official page names models and providers including Claude, GPT Image, DALL-E, Gemini, Grok, Seedream, Kling, Veo, Minimax, and Luma. It also describes 120+ tools, 93 guided workflows, and a searchable prompt library. 3 6

[editor's view] PromptReverse is the most expensive clone because the promise spans three systems: analyze the reference, keep the prompt useful, and route generation across models with different limits and output formats. The personal prompt library is the first information asset, but it becomes a moat only when saved prompts reliably improve the next project. A catalog of integrations alone is easy to copy and expensive to maintain.

[editor's view] Information edge: Three stars. Saved prompts, curated starter packs, and a user's history can make the product better over time. Legal risk: Four stars. The terms prohibit infringing uploads and scraping, while the privacy policy says uploads are sent to third-party AI providers and temporary files are deleted after processing. A clone needs clear rights handling for reference images and videos, plus a provider-by-provider policy review. 11 12

If you wanted to copy this

[editor's view] First concrete step: build one image-to-prompt workflow for one creator niche and test whether five users can reuse a saved prompt on a second project. First likely failure mode: the reverse-engineered prompt sounds detailed but fails to reproduce the reference, so the library becomes a scrapbook instead of a production tool.

What these three numbers actually say

[editor's view] The products sit at different points on the same wrapper spectrum. Bluff AI wraps generation in a content loop and has the strongest public revenue. Mascofast wraps generation in an identity-preservation workflow and has the smallest but clearest pilot. PromptReverse wraps several models in a reusable production system and has the biggest operating surface. That comparison is useful because it separates model access from the work around it.

[editor's view] The first validation metric should match the risk. For Bluff AI, measure consent and repeat sharing before adding templates. For Mascofast, measure whether the mascot stays recognizable across paid iterations. For PromptReverse, measure whether a saved prompt gets reused and produces a satisfactory second asset. The scores are a map of those tests, not a verdict about which idea you should build.

Sources

[data] The links below are the original product pages and first-party policy pages used for the claims above.

Source	Used for
[data] Bluff AI on TrustMRR	Revenue snapshot, founders, founding date, acquisition note, pricing, stack fields
[data] Bluff AI official site	Product flow and template examples
[data] Bluff AI privacy policy	AI provider, storage, analytics, and photo handling
[data] Bluff AI terms	Consent, deepfake, user-content, and output-use rules
[data] Mascofast on TrustMRR	Revenue snapshot, team, founding date, acquisition, pricing, and stack fields
[data] Mascofast official site	Product flow, Peerlist launch note, pricing, credits, and export claims
[data] Mascofast privacy policy	Retention, deletion, and training-use claims
[data] Mascofast terms	Third-party model, commercial rights, and output-risk terms
[data] PromptReverse on TrustMRR	Revenue snapshot, team, founding date, traffic, pricing, and stack fields
[data] PromptReverse official site	Tool catalog, model list, prompt library, and monetization
[data] PromptReverse privacy policy	Upload handling and third-party AI providers
[data] PromptReverse terms	Upload rights, acceptable use, and credit rules

What to do tomorrow morning

[editor's view] Pick the least risky workflow you understand, write one sentence naming the buyer's repeated job, and sell five manual trials before you build the wrapper. Track the first moment each tester hesitates: that observation will tell you whether your next problem is output quality, distribution, or trust.