# Boxer AI Aiolum story: https://aiolum.com/story/boxer-ai/ · added 2026-10-25 · updated 2026-10-25 **In short.** Boxer AI (AI boxing coach app that tracks punches with the phone camera, USA): US$5.4K MRR, October 2026. Evidence grade A: disclosed by a company or a filing, or verified by a payment processor. Team: 1 person according to the seller's TrustMRR profile; the seller is a student, named only by an X handle. Launched 8 March 2026; still running, revenue falling. AI boxing coach app that tracks punches with the phone camera · USA · AI-native · solo founder · still running, revenue falling · group: Health and appearance - Outcome: US$5.4K MRR, October 2026 (evidence grade of the headline number: A; headline in US dollars: 64716) - Launch: 8 March 2026; months from launch to the outcome: 7 - Team: 1 person according to the seller's TrustMRR profile; the seller is a student, named only by an X handle - Outside funding: None reported; listed as bootstrapped - Main factor: Niche - Meters, 1 to 4: evidence strength 3, copyable pattern 2, capital needed 2 - Tags: AI-native, Solo founder, Bootstrapped, Running, B2C, Health and fitness **The product.** An iPhone app that coaches boxing at home. The phone's camera tracks the user's jabs, crosses, hooks and uppercuts, and the app gives live coaching, video analysis, AR padwork and headwork, guided drills and workouts. After each session it shows a technique score, punch statistics, calories burned and personalized feedback. Users earn XP, climb levels and fight 1v1 matches against other users. The coaching, analysis, drills, workouts and leaderboard need a subscription. **Why it worked.** It needs nothing but a phone: the camera replaces the sensors, the bag and the coach that earlier fitness products had to supply. The game layer of XP, levels and 1v1 matches gives users a reason to come back. The seller lists TikTok, TikTok Ads and Instagram as its marketing channels. Evidence grades: A = disclosed by a public company, a filing or a payment processor, or verified by one; B = two or more independent reports agree, or a journalist confirmed it; C = a founder claim that no one else verified; D = reported, not verified: one founder claim or a third-party estimate with no second source; calc = computed on this page from graded facts. ## Summary Boxer AI is an iPhone app that works as a boxing coach. The user puts the phone in front of them and trains; the phone's camera tracks each punch, and the app scores the technique and gives feedback during and after the session. This is computer vision, software that recognizes objects and movements in camera images. Most of the app sits behind a paywall, a screen that asks for a subscription before the feature can be used. The App Store seller is Impact Development LLC, a Wyoming company that also sells AI coaches for wrestling, golf and basketball and a makeup app. The app was released on 8 March 2026. TrustMRR reads its sales through an API key, a password that lets one program read another program's data, for RevenueCat, a service that records in-app subscription purchases for app makers. On 25 October 2026 it showed US$44,509 of revenue since launch and MRR, monthly recurring revenue, in which a yearly plan counts as one twelfth of its price, of US$5,393 from 755 active subscriptions. A second tracker, whatsthe.app, read the same account until 15 September 2026, when the MRR was US$7,748 from 764 subscriptions. The revenue figures are grade A, because both sites read them from the app's subscription records, not from screenshots. Everything else is less certain. The seller is named only as the X handle @itsbenmarlowe, describes himself as a student, enters a team of one person and no funding, and offers the app for sale at US$35,000 on TrustMRR and US$150,000 on whatsthe.app. Revenue peaked in August 2026 at US$10,155 and was falling in October. The page also shows earlier motion-tracking and coaching companies that failed. ## Pricing Boxer AI in-app purchases on the US App Store, read 25 October 2026 (grade C) | Plan | Per month, billed yearly (US$) | Per month, billed monthly (US$) | Limit | |---|---|---|---| | Boxer AI Pro (four prices) | | | US$8.99, US$9.99, US$29.99 and US$34.99; no source gives the length of each plan | | Yearly Discounted | 1.25 | | US$14.99, named as a yearly discount | | 3 Months Discounted | | | US$19.99 for 3 months, about US$6.66 a month | The app is free to download, and live AI coaching, full analysis, the arena, drills, workouts and the leaderboard need a subscription. Apple lists six in-app purchases with names and prices but not the length of each plan, and whatsthe.app shows the same six prices. The per-month values are calculated from the plan names. One App Store review of 16 July 2026 complains that the price appears only after the whole registration. ## How much revenue does Boxer AI make? The numbers, graded US$44,509 of RevenueCat-verified revenue in seven and a half months and US$5,393 of MRR on 25 October 2026, down from a peak in August | Fact | Value | Grade | Note | Sources | |---|---|---|---|---| | Monthly recurring revenue (MRR), 25 October 2026 | US$5,393 | A | TrustMRR snapshot, revenue synced on 25 October 2026 at 11:39 UTC through a RevenueCat API key. 755 active subscriptions. TrustMRR gives MRR growth of -25.2% over the last 30 days. | [1] | | Run rate from the October 2026 MRR | about US$64,716 a year | calc | Calculated: US$5,393 times 12. | [1] | | Total revenue, March to 25 October 2026 | US$44,509 | A | TrustMRR, all time. The eight monthly values add up to US$44,510; the difference of US$1 is rounding. | [1] | | Revenue by month, 2026 | March US$4,619; April US$3,874; May US$1,763; June US$4,358; July US$6,752; August US$10,155; September US$8,585; October to the 25th US$4,404 | A | TrustMRR monthly timeline. March is the month of release. Calculated: August was 50.4% above July, and September 15.5% below August. | [1] | | Revenue in the last 30 days, to 25 October 2026 | US$5,940 | A | TrustMRR, which gives growth of -33.5% against the 30 days before; calculated, those earlier 30 days brought about US$8,932. Last 3 months: US$24,262. Last 6 months: US$36,207. | [1] | | Daily revenue, late September and October 2026 | about US$307 a day on 26 to 30 September, about US$160 a day on 14 to 24 October | calc | Calculated from TrustMRR's daily table: 47.8% lower. | [1] | | whatsthe.app figures, 15 September 2026 | MRR of US$7,748, US$9,064 of revenue in 28 days, 764 active subscriptions, 4,187 active customers | A | Read by whatsthe.app through RevenueCat ("revenuecat-verified" in the page data), last updated on 15 September 2026 at 20:40 UTC. MRR growth over 28 days: 0.0%. Calculated: the MRR on TrustMRR on 25 October was 30.4% lower. | [3, 4] | | MRR per active subscription | about US$10.14 in September, about US$7.14 in October 2026 | calc | Calculated: US$7,748 divided by 764, and US$5,393 divided by 755. The number of subscriptions hardly changed, so the fall in MRR came mainly from a lower value per subscription. | [3, 1] | | Asking prices | US$35,000 on TrustMRR; US$150,000 on whatsthe.app | C | Set by the seller. Both listings were first made on 23 August 2026 and carry the same seller message. TrustMRR counts 17 offers received; whatsthe.app counts 1 inquiry. | [1, 3, 4] | | Asking price compared with revenue | 0.5 times on TrustMRR; 1.4 times on whatsthe.app | calc | Calculated: US$35,000 divided by 12 times the last 30 days of revenue (US$71,280) is 0.49, as TrustMRR shows; US$150,000 divided by whatsthe.app's annualized revenue of US$108,800 is 1.38. | [1, 3] | | Seller | the X handle @itsbenmarlowe; no full name in any source | C | whatsthe.app names this handle as the founder of Boxer AI and of the company's makeup app, and the listing's contact address is on the impactdevelopment.llc domain. TrustMRR names no founder. | [3, 5, 6] | | Seller on the App Store | Impact Development LLC | B | Apple's record. The company's terms describe it as a Wyoming limited liability company. Its developer page lists five apps. | [2, 7, 8] | | Team | 1 person | C | Entered by the seller on TrustMRR. The seller's message says that he has started a major project with a co-founder while continuing his studies; that project is not Boxer AI. | [1, 3] | | Outside funding | none reported; listed as bootstrapped | C | TrustMRR profile. No source mentions investors. | [1] | | Self-reported profit margin | 50% | C | On both listings; whatsthe.app marks it as self-reported. | [1, 3] | | Users | 670 in the seller's estimate; 755 active subscriptions verified | C | TrustMRR shows the estimate next to the verified count, and they do not agree. whatsthe.app counted 4,187 active customers on 15 September 2026, with 18.2% of them subscribed. | [1, 3] | | App Store ratings, 25 October 2026 | 2.1K ratings, average 4.7 of 5 | B | Apple's page. TrustMRR's App Store sync of 12 October 2026 gives 2,029 ratings in the US store. | [2, 1] | | Age rating | 13+ | B | Apple's page, with frequent contests, user-generated content and social media. The company's terms say that users aged 13 to 17 need a parent's or guardian's consent and supervision. | [2, 8] | | Release | 8 March 2026 | B | Version 1.0 in the App Store history. whatsthe.app gives the same worldwide release date and TrustMRR the same founded date. | [2, 3, 1] | | Time from release to the October 2026 figure | 7 months | calc | Calculated: March 2026 to October 2026. | [1, 2] | How to read this table: a grade A value was read by TrustMRR or whatsthe.app from the app's RevenueCat account through an API key. A grade B value is a record that Apple shows on the App Store. A grade C value was written by the seller, on a listing, on the App Store or in the company's terms. A "calc" value was computed on this page from the rows above it. ## When did Boxer AI launch, and how did it grow? Timeline From release to a sale listing and a falling MRR in seven and a half months. Each entry carries its grade. - **12 June 2025** (grade C [6, 9]): Impact Development LLC releases the makeup app Glamify AI; the listing text calls it Blush AI, the name on the App Store today. Its sale listing later calls the app pre-marketing. - **8 March 2026** (grade B [2]): Boxer AI version 1.0 is released on the App Store, sold by Impact Development LLC. Three more versions follow the next day. - **March 2026** (grade A [1]): US$4,619 of revenue in the first month. - **May 2026** (grade A [1]): US$1,763 of revenue, the lowest month, 61.8% below March. - **2 to 15 June 2026** (grade B [2]): Seven new versions in two weeks, from 1.0.9 to 1.1.5. Every release note says only "Improvements". - **July 2026** (grade A [1]): US$6,752 of revenue, 54.9% more than June. - **August 2026** (grade A [1]): US$10,155 of revenue, the best month so far. - **23 August 2026** (grade A [1, 4]): Listed for sale on TrustMRR and on whatsthe.app on the same day, with the same message: the seller no longer has time for the app. - **25 August 2026** (grade B [2]): Version 1.2.0, the latest version on 25 October 2026. - **15 September 2026** (grade A [3]): Last whatsthe.app update: MRR of US$7,748, 764 active subscriptions, US$9,064 in 28 days. - **September 2026** (grade A [1]): US$8,585 of revenue, 15.5% below August. - **25 October 2026** (grade A [1]): TrustMRR: MRR of US$5,393, 755 active subscriptions, US$44,509 in total, 17 offers for the US$35,000 listing. ## Why did Boxer AI work? ### The phone camera as the only equipment The App Store page says that the camera tracks jabs, crosses, hooks and uppercuts and that every session can be done at home with only the phone. Earlier products in this area needed sensor clothing, a game console camera or a human coach. Here the user needs no device, no gym and no partner, which also makes the first try cheap. ### A game around the training Users earn XP, level up, compete in 1v1 matches with boxers worldwide and join a community; Apple's age rating notes frequent contests. One review dated 30 June describes competitions with a cousin. These parts give a reason to open the app again, and the leaderboard and arena are behind the subscription. ### Subscription after onboarding The app is free to download, and the paywall appears after registration. A review of 16 July 2026 complains about exactly that. Apple lists six in-app purchases, two of them named as discounts, from US$8.99 to US$34.99. ### One format, several sports The same seller sells Wrestler AI, Golfer AI and Hooper AI, and their App Store pages repeat the structure: camera or video coaching, XP and levels, global rankings, a community, simple nutrition tools and a subscription for the main features. A different founder built the same kind of app for boxing, Brawler, which the App Store shows next to Boxer AI. ### Short videos as the channel Both listings name TikTok, TikTok Ads and Instagram as the marketing channels. No source gives the ad spend or the number of videos, and the seller writes that "the previous strategy left significant room for growth" in marketing and distribution. ## How it was built These details come from the seller-entered fields of the TrustMRR profile and the whatsthe.app listing (grade C) and from the App Store page (grade B). - Swift and SwiftUI for the iPhone app; Python with FastAPI for the backend, running on Google Cloud and Firebase. - OpenAI is named in the technology list. No source says which model is used or whether the punch tracking runs on the phone or on a server. - RevenueCat records the subscriptions; TrustMRR and whatsthe.app read the revenue from it through API keys. - 21 versions in the App Store history from 8 March to 25 August 2026; the release notes say only "Improvements". The app is 175.7 MB, needs iOS 17 or later and is offered in English and six other languages. - Apple's privacy label says "Data Not Collected"; Apple notes that it has not verified this. The terms page, written for the company's makeup app, says that third-party services such as OpenAI may process users' data. - The company's website is a three-page Google Sites site with a contact address, a privacy policy and the terms. Sources: [1, 3, 2, 8] ## Who tried this before Boxer AI, and why did they fail? Four companies in the Loot Drop database tried to coach exercise through motion tracking or remote coaching. The values below are from that database and are not independently verified. The failed startups come from Loot Drop (https://www.loot-drop.io/), a free database of dead startups, retold here in our own words; Loot Drop describes its entries as AI-assisted summaries of public sources, so they carry no evidence grade. ### Enflux (2016 to 2019; funding US$1M; cause: Competition) - Promise: Smart clothing with sensors that track movement in 3D and give real-time feedback on form. - Why it died: The clothing was costly to make and needed a complex supply chain, and Fitbit and Apple kept adding features to their wearables. - Record: https://www.loot-drop.io/startup/1089-enflux ### FitFu (2011 to 2014; funding US$300K; cause: Competition) - Promise: A phone app that detected exercises with the phone's motion sensors and turned them into a game with rewards. - Why it died: Users left after the novelty wore off, it had no strong way to make money, and phones were not yet capable of better tracking. Nike Training Club and Fitbit offered more. - Record: https://www.loot-drop.io/startup/1194-fitfu ### Zigfu (2011 to 2014; funding US$200K; cause: No Market Need) - Promise: Tools that let developers add motion sensing with Microsoft's Kinect camera to their applications. - Why it died: It depended on one device, and Microsoft moved away from Kinect. Motion detection that needs no special hardware made its offer less useful. - Record: https://www.loot-drop.io/startup/1602-zigfu ### Charge Running (2021 to 2024; funding US$5M; cause: Competition) - Promise: Live online running classes with real-time audio coaching, like a personal trainer at home. - Why it died: Users preferred flexible times to scheduled classes, and AI fitness apps offered personal coaching on demand. - Record: https://www.loot-drop.io/startup/895-charge-running ## What changed Enflux put sensors in clothing, Zigfu depended on the Kinect depth camera, and FitFu had only the motion sensors of a 2011 phone. Each needed special hardware or ran into the limits of the hardware of its time. Charge Running used people as coaches, at fixed times. All four raised money, from US$200,000 to US$5 million, and closed. Boxer AI uses only the ordinary camera of an iPhone and software that recognizes punches, so it has no hardware to make and no coaches to pay, and one person could build and run it. What did not change is FitFu's problem: keeping users after the first weeks. Boxer AI's revenue in the 30 days to 25 October 2026 was 41.5% below its best month, August, and the seller says that he no longer has time to market it. Larger apps for the same sport, such as Brawler, compete for the same users. ## Can one person copy Boxer AI now? **Verdict: The format, yes. This app as an investment, with care.** An AI coach for one sport that people train alone, with the phone camera as the only equipment, a game layer and a subscription, is a format that several small sellers now repeat. The tools are open to anyone. What a copier cannot take from this case is a growing business: Boxer AI's verified revenue peaked in August 2026 and the seller has moved to another project. 1. Pick a sport that people practise alone in front of a phone, where a camera can see the whole movement. 2. Make the camera tracking the core feature, and add XP, levels and matches against other users so that people return. 3. Charge a subscription, and show the price early: one user's main complaint was that it appeared only after registration. 4. Connect the RevenueCat account to a verification site such as TrustMRR, so that a buyer can see real figures if you sell later. **What is still open.** Sports and exercises with clear, countable movements that people train at home, as long as no app with large ad budgets serves them yet. ## What could go wrong for Boxer AI? - **Falling revenue.** Revenue rose from US$1,763 in May to US$10,155 in August 2026, then fell: US$8,585 in September and US$5,940 in the 30 days to 25 October, 33.5% below the 30 days before. Daily revenue on 14 to 24 October, about US$160 a day, was 47.8% below the level of late September. The MRR fell from US$7,748 on 15 September to US$5,393 on 25 October while the number of subscriptions stayed near 760. - **Two listings, two prices.** The same seller message appears on TrustMRR at US$35,000 and on whatsthe.app at US$150,000, both listed on 23 August 2026. The whatsthe.app figures have not changed since 15 September 2026, although the site says that data is updated hourly and the page data does not mark the API key as expired, so its US$150,000 price rests on older and higher numbers. - **A seller known only by a handle.** No source gives the seller's full name. whatsthe.app ties the handle @itsbenmarlowe to Boxer AI and to the company's makeup app, and the listing's contact address is on the company's domain. The App Store names only the company, Impact Development LLC. The seller says that he is a student and now works on another project with a co-founder. - **Teenage users and camera data.** The app is rated 13+ and includes contests, user-generated content and social features. The privacy label says that no data is collected, which Apple has not verified, for an app that films users through the camera. The terms ask for a parent's consent for users aged 13 to 17. - **Dependence on OpenAI and on ads.** The seller names OpenAI in the technology list and TikTok Ads among the channels. A price change at the model provider or on the ad platform changes the cost of each user. No source gives the cost of either. ## Sources 1. TrustMRR profile of Boxer AI, revenue synced 25 October 2026 — https://trustmrr.com/startup/boxer-ai.md (grade A). Read the markdown profile. The revenue, MRR, subscription, monthly and daily figures come from its Verification Sources section, read through a RevenueCat API key. The founded date, team size, funding, tech stack, channels, user estimate and profit margin are entered by the seller; the sale data is TrustMRR's listing record. 2. Boxer AI on the US App Store, read 25 October 2026 — https://apps.apple.com/us/app/id6759840821 (grade B). Read in full: description, in-app purchases, version history, privacy label, age rating, seller, ratings and reviews. Seller, dates, ratings and age rating are Apple's records; the description and prices are written by the company. 3. whatsthe.app listing: Boxer AI, read 25 October 2026 — https://www.whatsthe.app/boxerai2026 (grade A). Read in full, including the data embedded in the page. The revenue metrics were read by whatsthe.app through RevenueCat, last updated on 15 September 2026; the asking price, seller message, channels, technology, profit margin and founder handle are entered by the seller. 4. whatsthe.app leaderboard, read 25 October 2026 — https://www.whatsthe.app/leaderboard (grade A). Read the listing data embedded in the page for the Boxer AI entry: founder field, verification type and creation date. 5. whatsthe.app founder page: @itsbenmarlowe, read 25 October 2026 — https://www.whatsthe.app/founder/itsbenmarlowe (grade C). Read in full. Lists two apps under this handle: Boxer AI and Glamify AI - Makeup & Beauty. 6. whatsthe.app listing: Glamify AI - Makeup & Beauty, read 25 October 2026 — https://www.whatsthe.app/llc.impactdevelopment.glamifyai (grade C). Read in full. The seller's text for the company's makeup app: release date, team size 1, bootstrapped, no marketing, asking price US$600. Read for the purpose check of the seller's other apps. 7. App Store developer page: Impact Development LLC, read 25 October 2026 — https://apps.apple.com/us/developer/impact-development-llc/id1816085564 (grade B). Read in full. Lists five apps: Golfer AI, Hooper AI, Wrestler AI, Boxer AI and Blush AI. 8. Impact Development LLC: Terms & Conditions — https://www.impactdevelopment.llc/terms-conditions (grade C). Read in full. Linked from the Boxer AI App Store page. Source for the company form, the age rules and the use of third-party AI services. The site's robots.txt returns a 404 page, so no rule blocks it. 9. Blush AI: Makeup & Beauty on the US App Store, read 25 October 2026 — https://apps.apple.com/us/app/blush-ai-makeup-beauty/id6746294667 (grade B). Read the description, age rating and seller, for the purpose check: the user previews a makeup look on a selfie of their own face. 10. Wrestler AI on the US App Store, read 25 October 2026 — https://apps.apple.com/us/app/wrestler-ai/id6801542604 (grade B). Read the description, age rating and seller, to compare the structure with Boxer AI. 11. Golfer AI: Swing Coach on the US App Store, read 25 October 2026 — https://apps.apple.com/us/app/golfer-ai-swing-coach/id6812569480 (grade B). Read the description, age rating and seller, to compare the structure with Boxer AI. 12. Hooper AI on the US App Store, read 25 October 2026 — https://apps.apple.com/us/app/hooper-ai/id6807834578 (grade B). Read the description, age rating and seller, to compare the structure with Boxer AI. 13. Loot Drop database export, 1,749 failed startups — https://www.loot-drop.io/ (grade C). Local export. AI-written summaries, not independently verified. Source for the failed predecessors.