a0.dev Thinks You Should Build a Mobile App Before Lunch

The Macro: The No-Code Dream Finally Has a Brain

The promise of “anyone can build an app” has been floating around since at least 2012. Bubble, Adalo, FlutterFlow, Glide. The list is long. The results have been mixed. Most no-code tools produce apps that feel like no-code apps. They work for internal dashboards and MVPs that never graduate past the prototype stage. The ceiling is real, and experienced developers can spot a no-code app in about three seconds.

AI code generation is changing the math. Instead of dragging and dropping pre-built components into a visual editor, you describe what you want and an AI writes actual code. React Native code, specifically, in a0.dev’s case. That’s a meaningful difference. React Native is a real framework used by real companies. Instagram, Shopify, and Discord all use it. An AI that writes competent React Native isn’t producing toy apps. It’s producing apps that could, in theory, sit next to hand-coded ones on the App Store and hold their own.

The market for AI-assisted development tools is getting crowded fast. Cursor, Replit, Bolt, and Lovable are all playing in adjacent spaces. GitHub Copilot handles code completion at the line level. What most of these tools share is a focus on web development. Mobile has been harder to crack because the deployment pipeline is genuinely more complex. You need to deal with App Store review, provisioning profiles, device-specific behavior, and the general overhead of shipping to iOS and Android simultaneously. Anyone who has published an app to both stores knows the process is roughly 40% coding and 60% bureaucracy.

The Micro: From Prompt to App Store in One Pipeline

a0.dev gives you an AI coding agent that writes and edits React Native code in real time. You describe features, it generates them. You want a database? Pick Convex or Supabase and it wires everything up. Need AI inference or image generation inside your app? Built-in APIs handle that. The pitch is that you go from idea to published app without ever opening Xcode or Android Studio.

The numbers they’re putting up are hard to ignore. Over 200,000 users have built more than 300,000 apps on the platform. Those are large numbers for a company that’s still in its YC Winter 2025 batch. The pricing tiers run from $20 a month to $800, scaling by usage. The top tiers are clearly aimed at people building real businesses on top of the platform, not hobbyists experimenting on a Saturday afternoon.

Seth Setse and Ayomide Omolewa cofounded the company. Seth is a CMU grad and serial app developer. Ayomide’s resume reads like a tour of technical heavyweights: Nvidia, Google, Paramount, Lockheed Martin. They’re a three-person team in San Francisco, which makes the scale of the platform more impressive. Three people running infrastructure for 200,000 users building 300,000 apps is either extremely efficient engineering or a ticking time bomb. I’m guessing it’s the former, given the backgrounds.

The one-click deployment to App Store and Google Play is the feature that matters most. Plenty of tools help you write code. Very few handle the entire pipeline from generation through build, store listing, and submission. If that actually works reliably, it removes the single biggest friction point in mobile development. Not writing the code. Getting the code onto phones.

Monetization tools are baked in too. Payment processing, subscription setup, revenue analytics. This is the kind of full-stack approach that makes sense for the target user. If you’re a solo creator using AI to build an app, you probably also want AI-assisted monetization. Having to integrate Stripe and build your own analytics dashboard defeats the purpose of the tool.

The Verdict

I think a0.dev is solving the right problem in the right order. Mobile app development has too many steps that have nothing to do with the actual product. Build configurations, signing certificates, store metadata, screenshot requirements. It’s tedious work that experienced developers tolerate and new developers abandon. Automating that pipeline end-to-end is a real unlock.

The risk is quality. AI-generated code works great until it doesn’t. Edge cases, performance optimization, accessibility compliance. These are the things that separate a demo from a product. At 300,000 apps, most of them are probably simple. The test will be what happens when users try to build something with complex navigation, offline support, or heavy state management. That’s where generated code tends to fall apart.

The competitive threat is also worth thinking about. Cursor and Replit are not going to ignore mobile forever. When they move into React Native or Flutter generation, a0.dev will need to have a moat beyond “we were first.” The deployment pipeline could be that moat. If they own the fastest path from idea to published app, the code generation part becomes less important than the surrounding infrastructure.

At 30 days, I’d want to see how many of those 300,000 apps are actually live on the stores versus abandoned in draft. At 60 days, retention by pricing tier. Are the $800/month users sticking? At 90 days, the question is whether the apps people build on a0.dev are getting real users or just sitting in the store at zero downloads. The platform is impressive. The output quality will determine whether it’s a tool or a toy.