Google Opens Native Android App Creation to Anyone With a Prompt

Google flipped the switch on native Android app creation inside AI Studio on May 19, letting users describe an app in plain language and get a working Kotlin build with Jetpack Compose. The web-based tool handles the boilerplate, sets up the project structure, and renders a live preview through an embedded Android emulator. Users can then install the app on a physical device by connecting it to their computer, or push it toward the Google Play Console for broader testing. Google had teased vibe coding in AI Studio back in October 2025, but this release bolts on full native Android output rather than web-only prototypes. According to The Verge, the whole loop from prompt to installable APK now takes minutes instead of the weeks of setup and coding that traditional Android development demands.

The process starts with a text prompt describing what you want the app to do. Gemini models inside AI Studio generate the Kotlin code, wire up the UI with Jetpack Compose, and spin up the emulator preview automatically. When you are ready to move beyond the browser, you connect an Android phone and push the build straight to it. The system also supports sharing builds with others for testing, which matters for anyone who wants feedback before committing to a Play Store submission. TechCrunch reports that Google is threading this into a broader push to let consumers find AI-built apps through Gemini on the Play Store and the web, suggesting these tools are not just for hobbyists but part of a longer play to flood the ecosystem with AI-generated software.

Apple has taken a noticeably cooler stance on the same trend. The Information reported that Apple is cracking down on vibe-coded apps, tightening review processes and raising barriers for software that is largely AI-generated. Google's move heads in the opposite direction, lowering friction and openly marketing AI Studio as a path from idea to published app. This divergence maps onto each company's broader platform philosophy. Apple has historically guarded the App Store gate to maintain quality and security; Google is betting that volume and openness will win, especially on Android where sideloading and alternative stores already loosen its control. The timing is not subtle. Apple's developer conference, WWDC, is weeks away, and Google is clearly eager to own the narrative around AI-native development before Apple says anything.

Professional Android developers are not being ignored here, though the messaging is split. On the same day the AI Studio news dropped, Google's Android developer blog published tips for using Gemini inside Android Studio, the professional IDE, including agent mode enhancements that came with the Otter 3 Feature Drop in January. The blog post gathered advice from Google engineers and Developer Experts on how to use AI assistance without surrendering code quality. That dual track, hobbyist tools in AI Studio and power tools in Android Studio, suggests Google wants to capture the full funnel from first-time builder to senior engineer. It also creates tension. If anyone can generate a working app in minutes, the value proposition for hiring specialized Android developers shifts, at least for simpler projects.

Getting an app onto a phone is one thing; getting it onto the Play Store is another. Adalo's guide for vibe-coded apps notes that builders still need to convert their prototype into an Android App Bundle, set up a Play Console account, meet Google policy requirements, and pass review. AI Studio does not eliminate those steps, though it may ease them. The broader question is how Google will handle review volume and quality if AI Studio succeeds in flooding the Play Store with generated apps. Google has not announced any special review track for AI-built software, and its existing policies around malware, deceptive behavior, and intellectual property still apply. For now, the path from prompt to published app exists but remains dotted with manual checkpoints that keep human judgment in the loop.

Google is signaling that this is not a one-off feature but a structural bet. The company is expanding Gemini-powered app discovery on the Play Store and web, which only makes sense if it expects far more AI-generated apps to exist. Research published around the same time, including work on XR Blocks and Gemini, shows vibe coding spreading beyond mobile into extended reality, suggesting the underlying pattern, natural language to working software, is being applied across Google surfaces. The open question is whether quality control can keep pace with creation speed. If AI Studio produces apps that crash, violate policies, or simply clutter search results, Google will face the same curation headache that has made Apple wary. For now, the company is choosing speed and openness over caution, and the next few quarters of Play Store data will tell whether that gamble pays off.