Key takeaways
- Google AI Studio's Android output is client-side Kotlin and Jetpack Compose; Firebase, secrets management and multiplayer remain web-only features.
- AI Studio publishes to internal testing only – those testers don't count toward the 12-tester, 14-day closed test that new personal accounts need before production.
- Never ship a Gemini API key inside an APK or AAB: route requests through your own server or through Firebase AI Logic with App Check.
- Google Play's AI-Generated Content policy requires in-app reporting or flagging of offensive AI output, without leaving the app.
- WebViewGold, made by our team, wraps the full-stack web app in an Android Studio project with Play Billing, App Links and push; appsubmitter.io handles signing, Data safety and the release.
Our recommended solution
Turn your Google AI Studio app into a real Android app with WebViewGold
WebViewGold wraps the Google AI Studio web app you already have in a native Android Studio project and adds push notifications, a native offline screen, deep links, a QR scanner and more – the native value reviewers at Google Play look for. Built by our own team, and we submit WebViewGold apps all the time.
Can you convert Google AI Studio to an Android app?
Yes – and unlike on iOS, Google AI Studio gives you a head start. Build mode can generate a native Android app in Kotlin and Jetpack Compose and publish it to Google Play's internal testing track for up to 100 testers. To convert a Google AI Studio project into an Android app anyone can install, you take it through closed testing and a production release in Play Console yourself.
Which route fits depends on what you built. A native Android app from AI Studio needs a path from internal testing to production. A full-stack web app – React, Node.js, Firestore, Firebase sign-in – won't carry those features into the Android generator, because AI Studio's Android apps are client-side only, so wrapping the deployed web app is usually faster. This guide covers both routes and the Google Play rules that apply either way; our iOS guide for AI Studio apps covers the App Store.
What AI Studio's Android builder does – and where it stops
Google's documentation, last updated in August 2026, describes the Android mode in detail (AI Studio Android docs):
- Code: a standard Gradle project in Kotlin with Jetpack Compose and Material 3 – single activity, single module, Compose only, no NDK or native code, phones and tablets only.
- Preview: an emulator that runs in the cloud and streams to your browser. Camera and photo capture, NFC and Bluetooth, GPS and Google Play services aren't available there, so test those on a real phone – AI Studio installs the app over WebUSB.
- Publishing: under Settings > Publish, AI Studio signs the app with a managed keystore, creates the Play listing (or uploads a new version) and publishes to the internal testing track.
- Hand-off: "Production releases must be managed directly in the Google Play Console."
- Export: a ZIP download; GitHub export isn't available for Android projects yet.
- Scope: client-side only. Firebase integration, Google Workspace APIs, secrets management and multiplayer are web-only features.
That last point shapes everything else. A client-side app has no safe place for secrets, no Firestore accounts and no server to verify purchases – things a public AI app usually needs.
Want your Google AI Studio app in Google Play without the trial and error?
Book a free consultation call: we check your web app, tell you which native features it needs and what to fix before review. Or hand it off – our App Specialists set up signing and CI/CD, prepare the store listing and submit the app for you.
Native Kotlin app or wrapped web app? An honest comparison
| Question | AI Studio's Kotlin app | Web app wrapped with WebViewGold |
|---|---|---|
| Which AI Studio features come along? | UI and on-device logic; no Firebase, secrets or multiplayer | Everything the web app has: Node.js server, Firestore, Firebase Authentication, Stripe secrets |
| Where does the Gemini key live? | The Android docs don't say – inspect the code before release | On the server, where AI Studio already keeps it |
| Device features | Anything you build in Kotlin | Ready-made modules: push, scanners, biometrics, App Links, Play Billing |
| Shipping changes | Every change is a new release in Play Console | Web changes go live with a deploy; native changes need a release |
| iOS later | A separate Swift app | The same web app in WebViewGold's Xcode project |
| Effort to production | Android Studio work: backend, signing, target API, accounts | Configuration plus web-side fixes for mobile |
| Main Google Play risk | Broken functionality when server-dependent features are missing | Limited functionality when the wrapper adds nothing beyond the website |
A third option is a Trusted Web Activity: your web app runs full screen in the user's browser, and Digital Asset Links prove that app and site belong together. It suits apps that already work as a progressive web app. AI Studio doesn't document PWA support, so you would add the web manifest and service worker yourself, for example before generating the project with Bubblewrap or PWABuilder.
Route A: take the generated Kotlin app to a production release
- Download the project as a ZIP, open it in Android Studio and commit it to your own repository right away.
- Check how it reaches Gemini. Search the code for API keys and Gemini endpoints. A key compiled into an APK or AAB can be extracted, and every request made with it counts against your project's limits. Route requests through your own backend, or use Firebase AI Logic: its Kotlin SDK keeps the Gemini API key on the server and works with Firebase App Check against abuse (Firebase docs).
- Settle identity and signing. Android's documentation is explicit: updates must use the same application ID and signing certificate as the original release. AI Studio signs with a managed keystore, and its docs don't say whether you can export it. Open the Play app signing page of the app AI Studio created and note which certificates are registered. If the app uses Play App Signing, Play Console's upload key reset lets you switch to an upload key you control. Before you have users, starting a fresh app with your own application ID is the cheap alternative.
- Raise the target. Since August 31, 2026, new apps and updates must target Android 16 (API level 36), which brings enforced edge-to-edge layouts and predictive back – see the target API level guide.
- Add what client-side mode left out: sign-in through Credential Manager (passkeys, passwords, Sign in with Google), account deletion, a report button for AI output and server-side purchase verification.
- Build an Android App Bundle. New apps have had to publish as AABs since August 2021. Sign it with your upload key and roll it out to a closed testing track.
Route B: wrap the AI Studio web app with WebViewGold
If your app relies on Firestore, Firebase sign-in, Stripe or multiplayer, keep the web app as the product and give it a native Android shell. WebViewGold, built by our own team, is a ready-made Android Studio project: you point it at your deployed URL and enable native modules. For an AI Studio app, these matter most:
- Google login: Firebase's default "Sign in with Google" fails in a WebView, because Google blocks OAuth there with a
disallowed_useragenterror and recommends Custom Tabs on Android. WebViewGold'sgooglelogin://prefix opens the login in a Chrome Custom Tab and then returns users to your web app; native sign-in with Credential Manager is the other clean option. Test the round trip with your Firebase project – and never alter the user agent to slip past the block. - Purchases: Google Play Billing for credits and subscriptions (Extended license).
- Notifications: push via Firebase, OneSignal or Pushwoosh. From Android 13, users must grant
POST_NOTIFICATIONSfirst, so ask when a long generation starts rather than at launch. - Links: App Links verified through
/.well-known/assetlinks.jsonon your domain; with Play App Signing, list the app signing key's SHA-256 fingerprint. - Input for Gemini: the document scanner and the QR and barcode scanner for vision features.
- Resilience: an offline fallback with a reconnect option for dead zones and Cloud Run cold starts.
- Re-entry and ratings: biometric authentication and the native Play review dialog.
Test two Android details in every build: the back gesture should move through your app's screens instead of closing it, and photo upload fields must open a picker – a WebView only does that when WebChromeClient.onShowFileChooser is implemented. WebViewGold is a one-time purchase, and the separately sold Cloud Builder builds and uploads the app from your browser if you don't use Android Studio. No tool can promise an approval, but native features in the main journey answer Google's rules on WebView apps and limited functionality far better than a bare frame.
The shortcut from Google AI Studio to Google Play
WebViewGold: your Google AI Studio app, native on Android
Instead of building native features from scratch, start from WebViewGold. You get ready-made Xcode and Android Studio projects for your website, with native modules you switch on in the configuration:
- Push notificationsOneSignal or Firebase, for real, personal events
- Native offline screenNo browser error pages when the connection drops
- Native splash screenApp-like start instead of a loading web page
- Deep linksLinks open the right screen inside the app
- QR & barcode scannerDevice features the website alone can't offer
- In-app purchasesNative purchase flows where the store requires them
No tool guarantees an approval: Google still judges what your app offers. Use the native features in your main user journey and keep proving that you own the website – our App Specialists review your WebViewGold app before it goes to Google Play.
From internal testing to production: the Google Play gates
Whichever route you take, the app passes the same Play Console gates. AI Studio's internal test is useful for quick feedback, but it doesn't move your production timeline forward.
| Gate | What Google requires | Note for AI Studio apps |
|---|---|---|
| Developer account | A personal or organization account with a one-time registration fee; organizations need a D-U-N-S number | AI Studio's publish flow uses the same account |
| Closed test | Personal accounts created after November 13, 2023: at least 12 testers opted in for 14 consecutive days, then Apply for production | Internal testers don't count, and Google says a user in your internal test "must first opt out of the internal test and then opt in to the open or closed test" |
| Bundle and signing | AAB, Play App Signing, an upload key | Decide early which keystore owns the app (Route A, step 3) |
| Target API | API 36 for new apps and updates since August 31, 2026 | Check the generated targetSdk |
| Data safety | Required for every app; apps active only on internal testing tracks are exempt | An app that lived only in AI Studio's internal track may never have needed the form – it does now |
| Account deletion | An in-app option plus a web link entered in the Data safety form | Applies as soon as users can create accounts, Sign in with Google included |
| AI-generated content | In-app reporting or flagging of offensive output | Covers chatbots and image or video generators |
| Payments | Google Play Billing for digital goods and services; alternative billing programs exist in some regions | AI credits and premium models are digital |
Our guides go deeper on the 12 testers / 14 days requirement and on account deletion. Google also warns that some developer accounts and apps get extended reviews of up to seven days or longer in exceptional cases, so leave room before a launch date.
Data safety and AI content rules for a Gemini-powered app
Google Play counts data as collected when it leaves the device and as shared when it goes to a third party – unless that party is a service provider, defined as "an entity that processes user data on behalf of the developer and based on the developer's instructions". Map each AI feature before you fill in the form:
| AI feature | Data safety types to consider |
|---|---|
| Chat or text prompts | Messages: Other in-app messages ("instant messages or chat content"), or App activity: Other user-generated content |
| Photo analysis | Photos and videos |
| Voice prompts | Audio files: Voice or sound recordings |
| History saved in Firestore | The same types, collected and stored – and covered by your deletion process |
| Firebase sign-in | Personal info: Email address, User IDs |
Your Gemini tier matters here. Under the Gemini API terms, Google uses content sent through unpaid services to improve its products, and human reviewers may read it; with paid services it doesn't use prompts or responses for that. Data used for Google's own product development is hard to describe as processing on your behalf, so run a public app on a paid tier – which the terms require anyway for apps offered in the EEA, Switzerland or the UK. The Data safety guide explains how to audit SDKs and fill in the form.
Google Play's AI-Generated Content policy adds a feature requirement: apps that generate content using AI "must contain in-app user reporting or flagging features that allow users to report or flag offensive content to developers without needing to exit the app". Its examples include text-to-text chatbots where the chat is a central feature and images or videos generated from text, image or voice prompts. A flag icon on each answer that sends prompt, output and reason to your backend is a simple way to meet it – and tells you where your system instructions need work.
Let appsubmitter.io handle the Google Play release
The Play Console part is where AI Studio's automation ends: keys, tracks, forms and policy emails. appsubmitter.io takes it over. Our App Specialists set up the upload key and Play App Signing, build release AABs in a CI/CD pipeline with fastlane and GitHub Actions, GitLab CI, Bitrise or Codemagic, set up the closed testing track and opt-in link (you recruit the testers), prepare the store listing with screenshot guidance, fill in the Data safety form with the Gemini data flows and submit the app – including the communication with the Google Play review team. An AI-powered pre-submission check with a human review on top looks for a missing report button, deletion link or outdated target API before Google does.
The app stays in your own developer account, and it doesn't have to be finished when you book. Code changes – a backend for Gemini calls, a deletion flow – are quoted separately and upfront. Book the Google Play service or start with a free consultation call. Heading to the App Store next? Our Guideline 5.6 article for AI Studio apps covers the iOS pitfalls.
Checklist before you submit
- You picked a route: the generated Kotlin app (Route A) or the wrapped full-stack web app (Route B).
- No Gemini API key ships inside the APK or AAB; requests go through a server or Firebase AI Logic.
- You know which keystore signs the app and use an upload key you control with Play App Signing.
- The release is an Android App Bundle that targets API level 36.
- A closed test runs with at least 12 opted-in testers for 14 consecutive days – none of them still in the internal test.
- Google login opens in a Custom Tab or runs natively through Credential Manager, never inside the WebView.
- Users can report or flag offensive AI output without leaving the app.
- The Data safety form covers prompts, photos, voice and stored history, and production runs on a paid Gemini tier.
- Account deletion works in the app and through a public web link entered in the Data safety form.
- AI credits and subscriptions use Google Play Billing.
Frequently asked questions
Can Google AI Studio publish my app to Google Play production?
Do AI Studio's internal testers count toward the 12-tester requirement?
Do I need Android Studio to publish an AI Studio app on Google Play?
Is it safe to put my Gemini API key in the Android app?
How long does Google Play review take?
Can I use a Trusted Web Activity instead of a WebView wrapper?
I built my app in Firebase Studio. Does this guide still apply?
Recommended solution
From Google AI Studio to Google Play: WebViewGold + appsubmitter.io
- 1 Build with WebViewGold Turn your website into native iOS and Android projects.
- 2 Add native value Push, offline screen, deep links or scanning in your main flow.
- 3 We submit it Our App Specialists submit and talk to the review team.
WebViewGold is made by our team (jocapps GmbH). It is a tool, not a guarantee – approval decisions are made by Apple and Google.
Sources and further reading
- Google AI – Build Android apps in Google AI Studio
- Play Console Help – AI-Generated Content policy
- Play Console Help – App testing requirements for new personal developer accounts
- Play Console Help – Set up an open, closed or internal test
- Play Console Help – Use Play App Signing
- Play Console Help – Provide information for Google Play's Data safety section
- Firebase – Firebase AI Logic
- Google AI – Gemini API Additional Terms of Service
Store policies and third-party products change regularly – always check the current versions. This article is independent advice and not affiliated with or endorsed by Apple, Google or any other company or product mentioned; all trademarks belong to their owners. WebViewGold and appsubmitter.io are made by our team at jocapps GmbH.