Q1 App Store Submissions Jumped 84% as Vibe Coding Tools Took Over

The App Store just had its biggest quarter for new submissions in years, and Apple is openly pointing at AI coding tools as the reason.

The Information first reported the numbers, citing data from mobile analytics firm Sensor Tower. 9to5Mac picked up the story on April 6, and it has since been covered by AppleInsider, The Next Web, Gizmodo, and MacObserver.

The headline number: 235,800 apps were submitted to the App Store in the first quarter of 2026, an 84% increase compared to the same quarter in 2025.

The Numbers Behind the Surge

The surge looks even larger when you zoom out. Between 2016 and 2024, new app submissions to the App Store actually fell 46%. The store had been getting smaller, year after year, for nearly a decade. Then 2025 reversed the trend, growing roughly 30% to almost 600,000 new apps for the year. Q1 2026 picked up where 2025 left off and accelerated.

Sensor Tower’s Abraham Yousef put it plainly: “We’ve seen an explosive growth of new apps over the past year.”

To handle the volume, Apple has been processing more than 200,000 app submissions per week over the last 12 weeks. The company says 90% of submissions are still reviewed within 48 hours, with an average review time of 1.5 days. Apple has also pushed back against any suggestion that review times are getting longer, telling reporters that human review is still required for every submission, with AI tools assisting reviewers behind the scenes.

Apple Is Naming the Tools by Name

What makes the framing unusual is that Apple itself, in conversations with The Information, attributed the surge specifically to AI coding tools. The two examples cited most often are Anthropic’s Claude Code and OpenAI’s Codex, the two terminal-native agents that have come to define the high end of the vibe coding stack over the past year.

The other two names that keep showing up are mobile-first builders. Replit launched its iOS Mobile Apps feature in January 2026, which lets users describe an app idea in natural language, generate a React Native build via Expo, and preview it on an iPhone through TestFlight. CNBC, eWeek, and others covered the launch as a turning point for mobile prototyping. Anything is the other name. It does roughly the same thing for non-developers: you describe what you want, and the app builds it.

Both Replit and Anything are also the two apps Apple has been most aggressively enforcing against. Apple pulled Anything from the App Store on March 30. Anything briefly returned on April 3, ran a $5,000 weekend hackathon to celebrate, and was gone again by Monday. Replit has been blocked from publishing updates to its iOS app since January.

We covered the broader enforcement story when it broke last month: Apple Is Blocking Updates to Vibe Coding Apps on the App Store.

The Two Stories Are the Same Story

Read separately, the surge story and the crackdown story sound like they contradict each other. One says vibe coding is producing more apps than the App Store has seen in nearly a decade. The other says Apple is removing the tools that produce them.

Read together, they describe a single shift. The pipeline of who can ship an iOS app is widening fast, and Apple is trying to figure out which parts of that pipeline it wants to keep gated.

The 235,800 figure does not distinguish between apps built by professional developers using Claude Code or Codex from a Mac, apps built by hobbyists using Replit or Anything from a phone, and apps built by experienced teams who have simply gotten faster. All three are growing at once, and that is the part Sensor Tower’s data captures clearly.

The Harvard Gazette ran a piece this week that frames the same dynamic from the academic side. Karen Brennan, a Harvard Graduate School of Education professor, taught a six-week experimental vibe coding course in fall 2025 with 92 students, none of whom were required to have prior coding or AI experience. Her summary of the appeal and the limitation lands in roughly the same place the App Store data does. Vibe coding lowers the floor for who can create software. It does not raise the ceiling for what counts as good software, and it does not come with the accountability frameworks (reliability, safety, security, maintainability) that professional engineering has built up over decades.

What This Means for the Tools in Our Directory

Most of the vibe coding tools in the VibecodedThis directory are in the surge, in the crackdown, or both:

Claude Code and Codex are the two agents Apple is naming as drivers of the surge. Neither is an iOS app, so neither is in any direct enforcement risk. They mostly run from the terminal on a developer machine.
Replit is in the surge and the crackdown. The Mobile Apps feature is generating real submissions. The iOS client has been frozen on a January build.
Bolt, Lovable, and v0 are web-based and do not distribute through the App Store. Apps built with them can still face standard App Store review on their own merits, but the platforms themselves are not being touched by the current enforcement push.
Firebase Studio and Google Antigravity are in a different lane. They generate web and full-stack apps, not iPhone binaries, and Apple’s enforcement does not currently touch them.
VibeCode is a separate iOS app that has been in the same enforcement bucket as Replit since January.

The cleanest read is that the App Store surge is being driven by two distinct groups using different parts of the directory. Professional developers are shipping more apps faster using the terminal-native agents. New builders are shipping apps for the first time using the mobile-first tools. Apple seems comfortable with the first group and uncomfortable with the second one.

What to Watch Next

WWDC 2026 is the next obvious checkpoint. Apple has historically used WWDC to announce changes to App Review Guidelines, and the November 2025 update already added anti-copycat rules and AI data-sharing disclosures without specifically mentioning vibe coding. Whether the next round names it directly will tell us a lot about whether Apple intends to keep doing this case-by-case or codify a broader policy.

The other thing to watch is what happens to the 235,800 figure in Q2. If submissions keep climbing while Apple continues removing the tools at the front of the pipeline, the gap between the two trends will get hard to ignore. If submissions flatten or fall, that would suggest the enforcement has worked. Either outcome reshapes the story.

For now, the data is clear about one thing. More software is being shipped to iPhones than at any point since 2016, and the tools doing the shipping are the same ones in our directory.