# Recipe One Aiolum story: https://aiolum.com/story/recipe-one/ · added 2026-10-25 · updated 2026-10-25 **In short.** Recipe One (AI recipe keeper app with meal planner and grocery list, Israel): US$2.1K verified revenue in one month, September 2026. Evidence grade A: disclosed by a company or a filing, or verified by a payment processor. Team: 1 founder, Oren Haliva, the app's seller on the App Store and developer on Google Play; no other staff named in any source. Launched June 2025; still running, revenue falling. AI recipe keeper app with meal planner and grocery list · Israel · AI-native · solo founder · still running, revenue falling · group: Documents, data and slides - Outcome: US$2.1K verified revenue in one month, September 2026 (evidence grade of the headline number: A; headline in US dollars: 25632) - Launch: June 2025; months from launch to the outcome: 15 - Team: 1 founder, Oren Haliva, the app's seller on the App Store and developer on Google Play; no other staff named in any source - Outside funding: None reported - Main factor: Timing - Meters, 1 to 4: evidence strength 3, copyable pattern 3, capital needed 1 - Tags: AI-native, Solo founder, Bootstrapped, Running, B2C, Food and cooking **The product.** A phone app for home cooks. The user shares a link to a TikTok, Instagram or YouTube video, a food blog page, a photo of a cookbook page or a PDF, and the app writes out the title, ingredients and steps as a clean recipe in the user's own collection. An AI cook called Maya answers questions while the user cooks, and the app plans the week's meals and builds a grocery list from the saved recipes. **Why it worked.** Many recipes now live in short videos, where the ingredients are spoken or shown and never written down. A model that can watch a video and write out a structured recipe turned an old app category, the recipe keeper, into something that works for these videos. One developer built it on rented models and sells it through the app stores as a subscription. 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 Recipe One is a recipe keeper: an app that stores the recipes a person wants to cook again, in one searchable place. Its difference from older recipe keepers is the import step. The user shares a TikTok, Instagram or YouTube video, a web page, a photo or a PDF, and Google's Gemini models extract the recipe from it. For a video, the file itself goes to Gemini, which reads it directly, so the app can write out a recipe even when the creator never typed the ingredients. Around the import, the app adds cookbooks, search, a cooking mode that keeps the screen on, a meal planner, a grocery list and an AI chat assistant called Maya. The App Store lists the seller as Oren Haliva, and Google Play lists him as the developer, with a contact in Israel; the revenue profile names the founder as Oren. A copy of the website from 28 June 2025 already offered to extract recipes from social media and videos. Revenue is shown on TrustMRR, a site that reads a startup's sales directly from its payment provider through an API key, so the founder cannot type the figures in. Here the provider is Superwall, a service that runs the subscription screens of phone apps and records each purchase. TrustMRR reports verified revenue of US$2,136 in September 2026 and US$24,106 in total by 25 October 2026. Six months reached US$2,000 or more, with the best month, US$3,005, in January 2026. This page grades every number by its evidence. The revenue figures are grade A because they come from payment records. The founding month, the user counts and the plan prices come from the founder, the company's pages and the app stores. The page also shows four failed startups that tried to save recipes and web content or to sell home cooking, and what changed since then. ## Pricing Recipe One prices in October 2026, from the in-app purchase list on the App Store page and the company's affiliate page (grade C); the website has no pricing page | Plan | Per month, billed yearly (US$) | Per month, billed monthly (US$) | Limit | |---|---|---|---| | Free | 0 | 0 | a limited number of free recipe imports | | Pro Annual | 3.33 | | US$39.99 a year; also listed at 19.99, 29.99 and 49.99 | | Pro Monthly | | 9.99 | also listed at 7.99 | | Unlimited | 29.99, period not stated | | intro offer listed at 19.99 | The app is free to start, with a limited number of free recipe imports; the paid plan is called Recipe One Pro. The App Store lists several prices for the same plan names (Pro Annual at US$19.99, US$29.99, US$39.99 and US$49.99; Pro Monthly at US$7.99 and US$9.99); the store does not say which price applies to whom. The affiliate page uses US$39.99 a year as its example. An Unlimited purchase is listed at US$29.99, with an introductory offer at US$19.99, and its billing period is not stated. Payments go through Apple and Google. ## How much revenue does Recipe One make? The numbers, graded A one-person recipe app that reached US$2,136 of verified revenue in September 2026 and US$24,106 in total, read from its subscription provider | Fact | Value | Grade | Note | Sources | |---|---|---|---|---| | Verified revenue, September 2026 | US$2,136 | A | TrustMRR monthly revenue timeline, read from Superwall through an API key. Last sync 25 October 2026 at 16:01 UTC. The profile does not say whether Apple's and Google's commission is already deducted. | [1] | | Verified revenue, all time to 25 October 2026 | US$24,106 | A | Just below US$25,000. The story qualifies because six months reached US$2,000 or more. | [1] | | Months at or above US$2,000 | 6: January, February, March, June, July and September 2026 | A | US$3,005, US$2,619, US$2,783, US$2,356, US$2,190 and US$2,136. | [1] | | Best month | US$3,005, January 2026 | A | Up from US$1,018 in November 2025 and US$1,153 in December 2025. | [1] | | Average month, April to September 2026 | about US$1,971 | calc | Calculated: (1,981 + 1,571 + 2,356 + 2,190 + 1,592 + 2,136) / 6 = 1,971. About 30% below the January to March 2026 average of about US$2,802. | [1] | | Verified revenue, last 30 days to 25 October 2026 | US$1,680 | A | The profile shows -22.6% against the 30 days before. October 2026 to the 25th: US$1,378. | [1] | | Monthly recurring revenue, 25 October 2026 | US$1,007 | A | Monthly recurring revenue (MRR) is the subscription money that comes in each month, with yearly plans spread over 12 months. The profile shows +14.6% in the last 30 days. | [1] | | Active subscriptions, 25 October 2026 | 357 | A | | [1] | | MRR per active subscription | about US$2.82 a month | calc | Calculated: 1,007 / 357 = 2.82. Close to a US$39.99 yearly plan (3.33 a month) with some cheaper prices, so most subscribers probably pay yearly; the profile does not show the plan mix. | [1, 2] | | Profit margin, last 30 days | 82.0% | C | Shown on the TrustMRR profile. The profile does not say how costs are counted or who entered them, so it is graded as a founder figure. | [1] | | Founded | 1 June 2025 | C | Founder-entered date on TrustMRR. A Wayback Machine copy of the website from 28 June 2025 already describes the product. | [1, 9] | | Users | 1,000+ home cooks, 100,000+ recipes saved | C | Company home page. TrustMRR shows 1,000 users, entered by the founder. Google Play shows 1K+ downloads. | [4, 1, 3] | | Store ratings, October 2026 | 4.6 from 144 ratings (App Store); 4.4 from 194 reviews (Google Play) | C | | [2, 3] | | Outside funding | none reported | C | No source mentions investors. The founder's only app on both stores is Recipe One. | [1, 10, 11] | | Time from launch to the September 2026 figure | 15 months | calc | Calculated: June 2025 to September 2026. | [1] | How to read this table: a grade A value was read by TrustMRR from the app's payment records through Superwall, the subscription provider. A grade C value comes from the founder, the company's own pages or the app stores. A "calc" value was computed on this page from the rows above it. ## When did Recipe One launch, and how did it grow? Timeline From a website that promised to extract recipes from videos in June 2025 to six months at US$2,000 or more by September 2026. Each entry carries its grade. - **1 June 2025** (grade C [1]): Founding date that the founder entered on TrustMRR. - **28 June 2025** (grade C [9]): The earliest Wayback Machine copy of recipeone.app: "Never lose a recipe with Recipe One. Extract recipes from any social media, videos, images, or text, and instantly get a structured, organized recipe for your recipes library." - **November 2025** (grade A [1, 2]): First month in the verified timeline: US$1,018. Version 1.3.2 adds a banner that shows how many free recipes are left. - **December 2025** (grade A [1, 2]): US$1,153. Version 1.4.2 adds accounts, sync between devices and sign-in with Apple and Google. - **January 2026** (grade A [1]): Best month so far: US$3,005. - **March 2026** (grade A [1, 2]): US$2,783. Version 1.5.0 translates recipes into 18 languages and converts units automatically; version 1.5.1 adds iPad support. - **April and May 2026** (grade A [1, 2]): Revenue falls to US$1,981 and US$1,571. Version 1.7.2 improves imports from websites, Pinterest and social pages. - **June and July 2026** (grade A [1]): US$2,356 and US$2,190. - **18 September 2026** (grade A [1, 2]): Version 2.0 adds a meal planner, a grocery list sorted by supermarket aisle and an AI chef. September revenue: US$2,136. - **15 October 2026** (grade C [2, 6]): Version 2.0.2 names the AI chef Maya; she remembers the user's diet and allergies. The terms of service, updated the same day, say the service does not give medical, nutritional or dietary advice. - **25 October 2026** (grade A [1]): TrustMRR: US$24,106 in total, US$1,680 in the last 30 days, MRR US$1,007 and 357 active subscriptions. ## Why did Recipe One work? ### Video became readable A recipe in a short video is spoken or shown on screen, not written. The privacy policy says the app uploads the video of a social media post to Google's Gemini Files API, and that Gemini reads YouTube videos directly. A model that reads video is what lets one developer offer import from TikTok and Instagram, the places where the home page says recipes hide. ### An old category with a new import step Recipe keepers are an old app category, and the App Store page lists ten similar apps next to this one. The organizing features (cookbooks, search, cooking mode, PDF export) are common to these apps. The import from videos, photos and scanned pages is the part that needs a model. ### Subscriptions through the app stores The app is free to start and sells Pro subscriptions through Apple and Google, with Superwall running the subscription screens and recording each purchase. TrustMRR read those records: 357 active subscriptions and US$1,007 of monthly recurring revenue on 25 October 2026. ### Rented services instead of staff The privacy policy names the parts: Gemini for the AI, Firebase and Google Cloud for sign-in, the database and files, Apify to fetch social media posts, Superwall for subscriptions, Mixpanel for analytics, AppStack for ad attribution and Expo for notifications. Each one replaces work a team would otherwise do. ## How it was built These details come from the company's privacy policy, the App Store version history and the founder-entered technology list on TrustMRR. They are grade C. - App: built with Expo, a framework for writing one React Native app for both iPhone and Android (TrustMRR technology list). The App Store page lists English and 18 more languages. - AI: Google's Gemini models through the Gemini API extract recipes from links, videos, photos and PDFs, translate recipes and estimate nutrition; the AI chef runs through Firebase AI Logic. The privacy policy says user content is not used to train models. - Fetching posts: Apify, a web-data service, retrieves the posts, captions and videos behind TikTok, Instagram, Facebook and Pinterest links that users save. - Back end: Firebase and Google Cloud for sign-in, database, file storage and servers. The website runs on Vercel, with PostHog and Vercel Analytics for website measurement and Airtable for a waitlist. - Money and growth: Superwall for paywalls and subscription management, Mixpanel for product analytics, and AppStack to see which ads lead to installs. The privacy policy says the company advertises on Apple Ads, Meta, Google and TikTok. - Other ways in: a Chrome extension that saves the recipe from the page in the browser, a connector that lets ChatGPT, Claude and Cursor read and edit the user's recipes, about ten free web tools (unit converters, a nutrition calculator, a recipe cleaner) and a private affiliate program for food creators that pays 30% to 50% of what referred users pay. Sources: [1, 5, 2, 7] ## Who tried this before Recipe One, and why did they fail? Four companies in the Loot Drop database tried to save recipes and web content in one place, or to make home cooking easier, before models could read a video. 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. ### Springpad (2008 to 2014; funding US$7.5M; cause: Unit Economics) - Promise: Collect, organize and find everything from web articles to recipes in one place. - Why it died: Users liked the free product, but the company could not find features that people would pay for, and Evernote moved faster. It ran out of money. - Record: https://www.loot-drop.io/startup/1893-springpad ### Kippt (2012 to 2017; funding US$400K; cause: Competition) - Promise: A bookmarking tool to save, organize and share web content. - Why it died: Pocket and Evernote offered more complete products, and its social features did not bring new users on their own. - Record: https://www.loot-drop.io/startup/278-kippt ### Delicious (2003 to 2017; funding US$2M; cause: Competition) - Promise: Save and tag web links from anywhere and share them. - Why it died: After Yahoo bought it, development stalled. It was slow to move to phones, where people started to save things inside each app. - Record: https://www.loot-drop.io/startup/2009-delicious ### Dinnr (2012 to 2013; funding US$120K; cause: Unit Economics) - Promise: Same-day delivery of pre-measured ingredients with a recipe. - Why it died: Same-day delivery needed a logistics network and stock that a small company could not pay for, and growth outside its first city made it worse. - Record: https://www.loot-drop.io/startup/1984-dinnr ## What changed Springpad, Kippt and Delicious saved links and web pages. They stored what the page already said, and when the content moved into apps and videos, a saved link stopped being useful. Springpad also shows the money problem: people liked a free place to save things, and few paid for it. Two things are different for Recipe One. First, a model can now read a video or a photo and write out the recipe, so the saved item is a usable recipe, not a link. Second, the app sells a subscription through the app stores from the first import, with a free limit, instead of hoping to charge later. Dinnr tried to solve home cooking with ingredients and delivery trucks; Recipe One is software only, so its costs are the rented model and services, and TrustMRR shows an 82.0% margin for the last 30 days. ## Can one person copy Recipe One now? **Verdict: Yes, the method is easy to copy, and many have.** The App Store page lists ten similar recipe-saving apps next to this one; one of them, Plum, describes itself as saving Instagram and YouTube videos. The models and fetching services that Recipe One uses can be rented by anyone. What one developer can still build is a better import for hard cases, a cleaner app and steady distribution. Revenue here also shows the limit of a crowded category: it peaked in January 2026 and has stayed near US$2,000 a month since April, according to TrustMRR. 1. Find an old app category where the hard step is reading messy input, such as a video, a photo or a scanned page. 2. Use a model that reads that input to do the hard step, and build the familiar app around it. 3. Charge through the app stores with a free limit, and use a subscription tool that records every purchase. 4. Add many small ways in: free web tools that rank in search, a browser extension, a connector for AI assistants and an affiliate program for creators. **What is still open.** The same import step can work for other content that lives in videos: workout routines, craft and sewing instructions, home repair steps, or language lessons. Each needs a structured format at the end, the way a recipe has ingredients and steps. ## What could go wrong for Recipe One? - **Revenue fell from its January peak.** TrustMRR shows US$3,005 in January 2026 and an average of about US$1,971 from April to September 2026, about 30% lower. The last 30 days to 25 October 2026 were US$1,680. Monthly recurring revenue rose 14.6% in the same 30 days, so the next months will show the direction. - **Gross or net is not stated.** The figures come from Superwall, which records app-store purchases. The profile does not say whether Apple's and Google's commission is already taken off, so the money the developer keeps may be lower than these figures. - **Other creators' recipes.** The app copies a recipe from another creator's video or page into the user's own collection. The app pages describe saving for personal use, sharing with friends and family, and PDF export; the website's sitemap lists 145 pages (tools, blog posts and legal pages) and no public recipe pages, so the app does not republish the recipes it imports. The terms tell users to check an imported recipe against its original source. - **Depends on other platforms.** Imports from TikTok, Instagram, Facebook and Pinterest go through Apify, and the AI through Google's Gemini. A change in those platforms' access rules or in Gemini's prices would hit the core feature. - **Affiliate page shows an income chart.** The affiliate page for food creators has a chart titled "Your Affiliate Income" with an axis up to 8k and a rising line, next to a worked example of US$240 a month for 1,000 readers at a 2% purchase rate and the words "Real payouts vary by country, plan, taxes, refunds, and localized pricing". It is a commission program for creators with an audience, not a promise to users. ## Sources 1. TrustMRR: Recipe One profile, Superwall-verified revenue (data as of 25 October 2026) — https://trustmrr.com/startup/recipe-one (grade A). Read the markdown profile (https://trustmrr.com/startup/recipe-one.md) in full, once, on 25 October 2026. Grade A for the metrics in its Verification Sources section (Superwall API key, not merchant of record, revenue last synced 25 October 2026 at 16:01 UTC). The description, founder name, founded date, user estimate and technology list are founder-written, grade C. 2. App Store: Recipe One: Recipe Keeper, by oren haliva — https://apps.apple.com/us/app/recipe-one/id6743317601 (grade C). Read in full on 25 October 2026. Source for the seller name, the description, the version history with dates, the ratings and the in-app purchase prices. 3. Google Play: Recipe One: Recipe Keeper, by Oren Haliva — https://play.google.com/store/apps/details?id=com.recipeapp.ai (grade C). Read in full on 25 October 2026. Source for the developer name and country, the download count and the rating. 4. Recipe One home page — https://www.recipeone.app/ (grade C). Read in full on 25 October 2026, including the questions and answers. Source for the product description, Maya and the user counts. 5. Recipe One privacy policy — https://www.recipeone.app/privacy (grade C). Read in full on 25 October 2026. Source for the AI models, the service providers, the Chrome extension, the AI assistant connector and the advertising platforms. 6. Recipe One terms of service, last updated 15 October 2026 — https://www.recipeone.app/tos (grade C). Read in full. Source for the AI accuracy section and the statement that the service gives no medical, nutritional or dietary advice. 7. Recipe One affiliate program — https://www.recipeone.app/affiliates (grade C). Read in full on 25 October 2026. Source for the commission tiers, the US$39.99 yearly price example and the income chart. 8. Recipe One: Just the Recipe, a free web tool — https://www.recipeone.app/just-the-recipe (grade C). Read in full. Shows that the web tool displays a cleaned recipe and saves nothing on the page. 9. Wayback Machine copy of recipeone.app, 28 June 2025 — https://web.archive.org/web/20250628115253/https://www.recipeone.app/ (grade C). Read in full. The earliest archived copy of the website; source for the first product description. 10. App Store developer page: oren haliva — https://apps.apple.com/us/developer/oren-haliva/id1802181552 (grade C). Read in full. Lists one app, Recipe One. 11. Google Play developer page: Oren Haliva — https://play.google.com/store/apps/developer?id=Oren+Haliva (grade C). Read in full. Lists one app, Recipe One. 12. 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.