Shipaton 2026 by the Numbers: What 3,336 Indie App Projects Show

RevenueCat's Shipaton ran from August 1 to September 30. When it closed, the Devpost gallery had 3,336 projects, and apps.shipaton.com, the showcase made by Perttu Lähteenlahti, listed 2,139 published apps from this year. The same showcase has 610 apps from 2025 and 333 from 2024.

That's a big enough sample to say something about indie mobile apps in 2026, so I went through all of it, every project page and every app page. This is the full version, with the complete category table and a section on how the numbers were counted.

Everything here comes from what's public: the Devpost project pages and the showcase. I don't have access to the submission portal, which is for the RevenueCat team and the judges, so anything builders shared only with them isn't counted, and the official numbers may differ.

Revenue figures are the ones builders wrote in their own Devpost submissions. I'm quoting them as written, and each one links to the write-up it came from. Counts that come from matching text, like app types or how many projects mention revenue, are approximate.

Shipaton 2026 at a glance

• 3,336 projects in the Devpost gallery, and 2,139 published 2026 apps in the showcase
• 88.6% of projects were built by one person
• Health & Fitness is the biggest category, with 304 apps
• Swift and SwiftUI is the most listed stack, in 38.2% of projects
• Claude is the most listed AI model, in 356 projects
• 18.4% of the published apps are on both the App Store and Google Play
• About 35 projects, roughly 1 in 100, say what their app earned

Almost everyone builds alone

2,955 of the 3,336 projects (88.6%) have one person on them. Another 276 are pairs. The biggest team in the event is TapLive, with 16 people.

Click to enlarge

More than 3,500 people are listed as builders. 228 of them entered more than one project, and 10 entered five or more.

About one solo builder in four already sells under a company name on the store listing, going by the seller names.

Health & Fitness is the biggest category now

Productivity was the largest category in the showcase in 2024 and in 2025. This year Health & Fitness passed it with 304 apps. Then come Productivity (288), Lifestyle (230), Games (204) and Education (185). Those five make up 57% of everything shipped, out of 55 categories.

Click to enlarge

Here are the ten biggest categories across all three years, with each one's share of that year's apps:

Games grew the most, from 3.4% of apps last year to 9.5% this year, nearly tripling their share. If you add the puzzle, casual, strategy and other game genres that are listed separately, there are 305 games, level with Health & Fitness.

Two smaller shifts show up in the table too. Education dropped from 11.5% to 8.6% of apps, and Food & Drink grew from 15 apps to 84.

What the apps actually do

Store categories are broad, so I also grouped the apps by what their name and tagline say they do. It's keyword matching, so the counts are approximate.

Click to enlarge

Screen time and focus blockers are the biggest group, at about 105 apps, just ahead of workout trackers at about 100. Then journals and mood trackers (about 84), sleep and meditation (80), habit trackers (72), budgeting (70), baby and parenting (66), pets (58), calories and nutrition (53) and language learning (44). 229 apps have "AI" in the name or tagline.

Some I keep thinking about:

• SignBridge, a five-person team from Taiwan, built two-way Taiwan Sign Language translation.
• 1% Lock is a screen time blocker with a social feed inside it. Every post is checked by AI before it shows up, and "a normal check costs under one US cent."
• ambient: Evolving Soundscapes by zoome LLC runs on iPhone, iPad, Mac, Android and Apple Watch, the only app in the showcase on five device types.
• layrd: Photo Layout Studio is the only Vision Pro app.

The most-liked project in the gallery is Blindspot, a game where you photograph cars you spot and build a collection, with 38 likes. Most projects got none: 70% of the gallery has zero likes.

What they built with

Native Swift and SwiftUI is the most common stack, in 38.2% of projects. Then React Native and Expo (27.7%), Flutter (16.6%), native Kotlin (14.4%), Capacitor and Ionic (5.4%), Kotlin Multiplatform (4.1%) and Unity (1.3%).

Click to enlarge

About 29% of projects list an AI model. Claude shows up most (356 projects), then Gemini (336) and OpenAI (272). 57 list Apple Intelligence or its on-device Foundation Models.

Then there are the AI tools used to write the code itself: 129 list Codex, 107 Claude Code and 38 Cursor. On the backend, Firebase (650) and Supabase (632) are almost tied.

Where they shipped

56.1% of the published apps (1,201) are on the App Store only, 25.5% (545) on Google Play only, and 18.4% (393) on both. Last year 20.2% were on both.

Click to enlarge

More than half the projects (55.5%) are built with a cross-platform framework, meaning React Native, Flutter, Capacitor, Kotlin Multiplatform or Unity, yet fewer than one in five apps shipped to both stores.

12 apps are on the Samsung Galaxy Store, and all 12 are from this year. 62 apps ship an Apple Watch version, while 142 write-ups mention the Watch.

What it earned

Very few builders said what their app earned. Only about 35 projects, roughly 1 in 100, give a revenue or MRR number for their own app, and 14 give their MRR.

Click to enlarge

The chart shows four of them. Here they are, plus the ones that reported totals:

• Buck, a coin identifier: "over $13K in revenue ($13k in the last 28 days) and is currently at $6K MRR."
• Mio, a self-care habit tracker: "6,200 installs, 800 subscriptions, and $30,000 in revenue."
• Scran launched on August 6 and has "172 paying subscribers / $5,136 in gross revenue / A 52% trial-to-paid conversion rate."
• Clucky, a loud alarm clock from a team of two: "7,259 new customers, 743 trials, 153 paid subscriptions and $1,748 revenue." It's now at $261 MRR.
• Ritmo got to $830 MRR in under four weeks.
• Mekuru went from $7 MRR on September 11 to $369 on September 29.
• Paternal lists "Revenue: $23,385" for its first month.

The biggest number belongs to macro, a calorie counter for the Philippines: "around 20,000 paying customers, $88K in revenue, and more than 75,000 users." Most of that came before Shipaton. The app grew through one large partner until that partnership ended in July. It sells a lifetime unlock for 299 pesos, about $5, and the builder writes that they're still working out how to balance a cheap lifetime product with recurring revenue, which is why they added optional monthly plans.

How they priced

This is the part I know best, since pricing is what I work on.

CalHabit ran ads and wrote down what happened: "Kazakhstan gave me 50 of my 74 paid installs at $0.67 each and produced no revenue; the US and UK gave me 7 installs at $3.05 and produced both of the purchases I could trace to ads."

1% Lock started with a low price, then saw conversion come in "at around 17%, much higher than I had planned for. So I tested it and more than doubled the price. Conversion fell, and I am still testing to find the right price."

Seven projects say they set prices by purchasing power, and a few more set lower prices for specific countries. Some examples:

• BookLore AI ran "a purchasing-power-parity pricing pass across ~170 App Store territories using real World Bank PPP and exchange-rate data."
• Litmark set prices "by purchasing power in four tiers, from full price down to about 65% off, and Indonesia set by hand."
• CareBridge priced its subscription at about $1.99 to $2.99 a month in Vietnam so rural families can afford it.

Seven out of 3,336 is a small number, and it lines up with what I see outside the hackathon too: most apps launch with the store's default conversion of one base price. If you want the background, I've written about what localized pricing is and what the stores do with your price by default.

How I counted

Some notes on method, so you know how far to trust each number.

• Sources. Project counts, team sizes, Built With entries, likes and revenue quotes come from the public Devpost gallery. Published apps, store categories, stores and device types come from apps.shipaton.com. Both were pulled after submissions closed on September 30 and re-checked on October 3.
• Two different totals. The 3,336 is every public Devpost project. The 2,139 is published 2026 apps listed in the showcase. Percentages about stacks and teams use the first number, and percentages about categories and stores use the second.
• Self-reported figures. Revenue, MRR and subscriber numbers are what builders wrote. I haven't verified any of them.
• Keyword matching. App types, the "AI in the name" count, revenue mentions and pricing approaches come from matching words in names, taglines and write-ups. They're approximate in both directions.
• Seller names. The "one in four solo builders sell under a company name" figure compares builder names to store seller names, so it's an estimate.
• Earlier years. The showcase's 2024 and 2025 data is less complete on platforms, so the year-over-year comparisons only use category counts and App Store or Google Play presence.
• Likes are a snapshot and will keep changing.

Browse them yourself

All the apps are on apps.shipaton.com, and the full project write-ups are in the Devpost gallery. Thanks to Perttu for building the showcase and to RevenueCat for running the event.

If you shipped during Shipaton and want to see what your price looks like in other countries, I made a free tool for that: the Price Localizer shows the full country grid for any base price, no login. When you want to push those prices to the App Store and Google Play, PricePush does it in one step, and you can start free.