PreLoved AI: Revenue, team and how it was built
PreLoved AI (iPhone app that writes a Vinted listing (title, description, hashtags) and makes an optional AI model photo from one photo of a clothing item, United Kingdom): US$3.6K monthly recurring revenue, November 2026. Evidence grade A: disclosed by a company or a filing, or verified by a payment processor. Team: 1 developer named, Daniel Steel (App Store provider and privacy policy, trading as Aurora Apps); headcount not published. Launched August 2025; still running, revenue growing.Added 6 November 2026
- Outcome
- US$3.6K monthly recurring revenue, November 2026Revenue growing
- Evidence
- A disclosed by a company or a filing, or verified by a payment processor
- Launch
- August 2025
- Status
- Still running
- Team
- 1 developer named, Daniel Steel (App Store provider and privacy policy, trading as Aurora Apps); headcount not published
- Outside funding
- None disclosed
- Main channel
- Inside another platform
PreLoved AI is a subscription app for sellers on Vinted. It uses a vision-language model, an AI model that takes an image as input and writes text about it, to turn a photo of a garment into a ready-to-paste listing. A second model makes a virtual model photo: a generated image that shows the same item worn by an invented person, or on a mannequin or hanger. The developer, Daniel Steel, trades as Aurora Apps in the United Kingdom; the TrustMRR profile says the app launched in August 2025.
TrustMRR, a site that reads revenue directly from a company's payment provider, connects to the app's RevenueCat account. RevenueCat is a service that tracks App Store subscriptions for app developers. On 6 November 2026 the profile showed US$3,578 in monthly recurring revenue (MRR, the subscription revenue that repeats each month), 403 active subscriptions, US$3,765 of revenue in the last 30 days and US$28,222 since the start. October 2026 was the best month so far, at US$4,028.
The figures are grade A because they come from the payment records, not from the founder. They cover iPhone purchases only. This page also lists open questions: generated model photos on a marketplace that shows real items, store counts that are larger than the paying base, and free access given in exchange for social posts.

How much revenue does PreLoved AI make?
Each row shows the evidence grade and the source numbers from the list at the end of this page. The grades are explained in How Aiolum verifies.
| Fact | Value | Grade | How we know | Sources |
|---|---|---|---|---|
| Monthly recurring revenue, 6 November 2026 | US$3,578 | A | TrustMRR metric snapshot; payment source RevenueCat (API key), last synced 6 November 2026. Growth over the last 30 days: +27.6%. | [1] |
| Revenue, last 30 days to 6 November 2026 | US$3,765 | A | TrustMRR, up 15.2% on the 30 days before. | [1] |
| Revenue since the start | US$28,222 | A | TrustMRR all-time figure. The monthly table starts in December 2025 at US$1,340. | [1] |
| Best month | US$4,028, October 2026 | A | TrustMRR monthly table. Earlier peak: US$2,700 in March 2026; low: US$1,951 in July 2026. | [1] |
| Active subscriptions, 6 November 2026 | 403 | A | TrustMRR metric snapshot. | [1] |
| Monthly recurring revenue per subscription | about US$8.88 | calc | Calculated: 3,578 divided by 403 equals 8.88. | [1] |
| Annual run rate | about US$42,936 | calc | Calculated: 3,578 times 12 equals 42,936. | [1] |
| Profit margin, last 30 days | 83.0% | A | TrustMRR snapshot. TrustMRR does not say which costs are included. | [1] |
| Launch | August 2025 | C | Founder-written TrustMRR description: launched in August 2025. The Aurora Apps projects page dates the PreLoved AI entry 2 August 2025. | [1, 6] |
| Team | 1 developer named; headcount not published | C | The App Store provider is Daniel Steel; the privacy policy names the operator as Aurora Apps / Daniel Steel. In-app messages mention the PreLoved AI team; no source gives a number of people. | [2, 5] |
| Outside funding | none disclosed | C | No source mentions investors. | [1, 5] |
| App Store rating, 6 November 2026 | 4.2 out of 5 from 29 ratings | C | UK App Store page; age rating 4+. | [2] |
| Sellers claimed in the store description | 30,000+ | C | App Store description; TrustMRR shows the same founder-written estimate. Not verified, and much larger than the 403 paying subscriptions. | [1, 2] |
| Markets | UK first, then Germany, the Netherlands and France | C | Founder-written TrustMRR description; the app interface has English and 7 other languages. | [1, 2] |
| Marketing channels | App Store optimization and TikTok | C | Founder-written TrustMRR profile. The privacy policy also mentions measuring Apple Search Ads campaigns. | [1, 5] |
| AI models | Google Gemini and Vertex AI; OpenAI as backup; Kie AI for model photos | C | Privacy policy updated 26 September 2026. | [5] |
| Exchange rate used | US$1.3398 per £1, 30 October 2026 | A | Federal Reserve H.10 release of 2 November 2026. | [7] |
| Time from launch to the headline | 15 months | calc | Calculated: August 2025 to November 2026. | [1] |
How to read this table: a grade A value was read by TrustMRR from the app's RevenueCat account, or is an official exchange rate. A grade C value comes from the developer: the App Store description, the product website, the privacy policy or the founder-written part of the TrustMRR profile. A "calc" value was computed on this page from the rows above it.
| Plan | Per month, billed yearly (US$) | Per month, billed monthly (US$) | Limit |
|---|---|---|---|
| Free | 0 | 1 AI listing a month | |
| Basic | 6.69 (£4.99) | 20 listings a month | |
| Standard (most popular) | 17.40 (£12.99) | 250 listings a month | |
| Pro | 24.10 (£17.99) | 500 listings a month | |
| Elite | 40.18 (£29.99) | 1,000 listings a month |
Apple bills the plans; they renew until cancelled. The free plan is one item a month. All paid plans include virtual model photos, batch listing, history and copy buttons. The same listing shows Basic Annual at £29.99 (about US$40.18) and Standard Annual at £99.99 (about US$133.97) a year, and special offers of £2.99 and £6.99, without a stated allowance next to them, so the yearly column is empty. The founder-written TrustMRR profile gives US dollar prices of US$4.99, US$12.99, US$19.99 and US$29.99; other countries set their own prices.
When did PreLoved AI launch, and how did it grow?
From launch in August 2025 to US$3,578 of monthly recurring revenue in November 2026. Each entry carries its grade.
- August 2025CPreLoved AI launches on the App Store as an AI assistant for Vinted sellers. [1, 6]
- December 2025AFirst month in the TrustMRR revenue table: US$1,340. [1]
- January 2026AVersion 1.9.15 adds person, mannequin and hanger modes and a faceless option for model photos. Revenue that month: US$2,250. [1, 2]
- March 2026ARevenue reaches US$2,700, the peak of the first half of the year, then falls to US$1,951 by July. [1]
- 24 June 2026CAfter a review criticised the ethnicity and age fields of the model-photo feature, the developer replies that the categories were changed and that the feature can be turned off. [2]
- 19 August 2026CVersion 1.9.30 retires the Vinted search alerts and the photo studio to focus on listings. [2, 3]
- October 2026ABest month so far: US$4,028. Version 1.9.39 follows on 17 October. [1, 2]
- 6 November 2026ATrustMRR shows US$3,578 in monthly recurring revenue and 403 active subscriptions. [1]
Why did PreLoved AI work?
Four causes, in the order they mattered. Each one is a reusable lesson.
One marketplace, one job
The app serves sellers on one marketplace, Vinted, and does the part they repeat for every item: the title, the description, the hashtags and a clean photo. A narrow tool is easy to explain in an App Store search result, and the founder lists App Store optimization as a main channel.
Photo in, listing out
A vision-language model reads the garment and its care label from the photo. An image model then generates the item on a virtual model or a hanger. The privacy policy names Google Gemini and Vertex AI as the main models and OpenAI as a backup, so the developer rents the AI and builds the camera flow, prompts and checks around it.
No login to the marketplace
The app does not post to Vinted and does not touch the seller's Vinted account. The user copies the text and pastes it in. This keeps the app outside Vinted's account system and avoids the automation that marketplaces often forbid.
Usage-based plans
Plans are sized by the number of listings a month, from 20 to 1,000, so occasional sellers pay little and resellers pay more. TrustMRR shows about US$8.88 of monthly recurring revenue per active subscription.
How it was built
These details come from the founder-written TrustMRR profile, the App Store page and the privacy policy. They are grade C. [1, 2, 5]
- A native iPhone and iPad app in SwiftUI, with a Next.js website, according to the TrustMRR tech list.
- Google Firebase and Google Cloud store accounts, listings, messages and photos, and run the server functions and push notifications.
- Google Gemini and Vertex AI write descriptions, make model photos and check their quality; OpenAI is a backup; Kie AI also makes model photos.
- Apple in-app purchases with RevenueCat for subscriptions; the privacy policy also mentions PayPal and older Stripe subscriptions for a web app.
- Google Analytics for Firebase for usage, Sentry for crash reports, and Apple Search Ads measured through RevenueCat.
- Frequent releases: the App Store history shows about 25 versions between January and October 2026, many of them about photo quality.
Tools named in the sources
Each one opens its page in the Tools tab, with every story that uses it.
Who tried this before PreLoved AI, and why did they fail?
Three companies in our database of failed startups tried to make second-hand selling easier before AI could write a listing from a photo. The values below are from that database and are not independently verified.
Tradesy
The promiseA marketplace where people resell their pre-owned designer clothing and accessories.
Why it diedThe RealReal and Poshmark grew faster and earned more trust, and Tradesy could not set itself apart while paying high costs to win buyers and sellers.
Poshmark UK
The promiseSocial selling of second-hand fashion between people in the UK.
Why it diedDepop and Vinted had stronger communities and mobile apps, and the service did not adapt to how UK buyers shop.
Selltag
The promiseList and sell second-hand goods to people nearby with less effort than classified ads.
Why it diedIt never reached the scale a marketplace needs; Wallapop and then Facebook Marketplace had more buyers and sellers.
The failed startups come from our database of 1,749 dead startups, retold here in our own words. Its entries are AI-assisted summaries of public sources, so these cards carry no evidence grade.
What changed
The earlier companies tried to own the marketplace itself, and they lost to larger marketplaces with more buyers. Tradesy and Poshmark UK raised more than US$150 million each and still closed. Making a listing easier was a feature inside their apps, not a business.
PreLoved AI does not compete with Vinted; it sells a small tool to Vinted's sellers. Two things made that possible for one developer: models that read a garment and its label from a photo, and image models that put the garment on a generated person. Both are rented through APIs, so the developer's work is the app, the prompts and the photo checks.
Can one person copy PreLoved AI now?
Yes, for one marketplace and one country at a time.
The method is plain: pick a marketplace where sellers write many similar listings, and build a phone app that writes them from a photo. The same idea exists for eBay and Etsy. The risk is that the marketplace adds the same feature itself, or changes its rules about generated images.
Where it is still open: Other second-hand marketplaces and languages are open: furniture, books, children's goods, or local marketplaces in countries where Vinted is small. A tool that also suggests a price from sold items, as some eBay apps do, is another step.
- 1Choose one marketplace and learn what its buyers search for and what its listing form asks.
- 2Rent a vision-language model for the text and an image model for photos; spend the work on the camera flow and quality checks.
- 3Never log in to the marketplace for the user; let the user paste the result.
- 4Price by the number of listings a month, so heavy sellers pay more.
What could go wrong for PreLoved AI?
The risks are the mirror image of the causes above.
Generated model photos on Vinted
The app shows the seller's item on an invented person, and the default had been a person photo. Buyers on Vinted may not know that such a photo is generated. Vinted's own rules on generated images could not be read for this page, because vinted.co.uk's robots.txt disallows ClaudeBot. The app tells users to paste the text and add the photos themselves, and offers flat-lay and hanger photos without a person. A June 2026 review criticised the ethnicity and age fields of the feature; the developer replied on 24 June 2026 that the categories were changed and that the feature can be turned off.
Counts and wording in the store
The App Store description says the app is "Trusted by 30,000+ Vinted sellers", and TrustMRR shows the same founder-written estimate. TrustMRR counts 403 active subscriptions; the larger number is not verified and may count free users. The description also ends with "Sell faster, list more, earn more" and speaks of resale side hustles, without any figure for what users earn.
Free access for social posts
The promo page offers a code for "free access" to people who post about the app on Instagram, TikTok, YouTube and other networks. The page read on 6 November 2026 does not say that such posts must be labelled as advertising.
Revenue covers iPhone only, and platform details disagree
TrustMRR reads RevenueCat, which covers App Store purchases. The privacy policy (26 September 2026) also describes a web app and an Android app paid through PayPal, and older Stripe subscriptions; the product website says there is no Android app. Revenue from those channels, if any, is not in the figures. The earlier Vinted search alerts were retired in version 1.9.30; the sources do not say how that feature got its data.
Competition
Vinted could add the same listing help itself, and similar apps on this site serve other marketplaces: Value AI writes eBay listings (about US$2,300 a month, grade C), ValorixAI writes Etsy listings (no revenue figure), and PureZap edits marketplace product photos (grade A, small revenue).
Dependence on model providers
The privacy policy says Google Gemini and Google Vertex AI make the descriptions and model photos, OpenAI is a backup, and Kie AI also makes model photos. Price changes, rule changes about generating images of people, or quality changes at these providers reach the app directly.
The owner's other apps
Aurora Apps also lists Expense Camera, Cipher, Air Quality Watch, Drop the Top! and, marked soon, Datum: Body Fat Calculator and other utilities. The PreLoved AI pages read here do not link to them.
Sources
The grade next to each source is the best grade that the source can support. "Read in full" means that the text of the page was downloaded and the numbers were checked in it.
- A TrustMRR profile: Preloved AI, synced 6 November 2026Read the markdown profile. Revenue, MRR, subscriptions and margin come from the app's RevenueCat account (grade A); the description, prices, user estimate, tech stack and channels are founder-written (grade C).
- C App Store (UK): PreLoved AI: Vinted Listings, by Daniel SteelRead in full: description, plans, in-app purchase prices, ratings, reviews with the developer's replies, and version history to 1.9.39.
- C PreLoved AI website: Vinted listings written from a photoRead in full, including the pricing table and the FAQ.
- C PreLoved AI website: Promo codesRead in full.
- C Aurora Apps: PreLoved AI privacy policy, last updated 26 September 2026Read the sections on the operator, payments, analytics and service providers.
- C Aurora Apps: ProjectsRead in full. Source for the PreLoved AI date and the owner's other apps.
- A Federal Reserve H.10 Foreign Exchange Rates, release of 2 November 2026Read the United Kingdom pound row: US$1.3398 per £1 on 30 October 2026. Used for the US dollar prices on this page.
- C Database of failed startups, local exportLocal export. AI-written summaries, not independently verified. Source for the failed predecessors.
Run this company? Send corrections or a verified figure to aiolum@livyden.com