On this page
On November 17, 2026 (with @GoogleAIStudio posting at 22:23:01 UTC, or November 18, 2026 at 06:23:01 Asia/Shanghai; @OfficialLoganK posting at 06:25:45 Asia/Shanghai; and the Google Developers Blog post dated November 17, 2026 without a separate timestamp), Google introduced two major capabilities in Google AI Studio Build: custom domain support for web apps and an end-to-end full-stack Android app workflow with personal Google Play library publishing.
These updates expand Google AI Studio's application-building environment rather than launching a new foundation model, and are separate from our earlier report on one-click Google Maps Platform integration in Google AI Studio (see also our Google AI Studio overview page). The blog post notes that developers use existing Gemini 4 Carbon and Hydrogen models in AI Studio to build these apps (showing generativeModel("gemini-4-hydrogen") in sample code), with the workspace entry point at https://aistudio.google.com/apps?features=build_android_app gated behind Google account sign-in.
Custom Domains in AI Studio Build
Per @GoogleAIStudio, developers can now turn an idea into a live, branded app with a custom domain directly inside AI Studio Build. In a companion recap post on X, Logan Kilpatrick (@OfficialLoganK) added that users can buy custom domains directly in AI Studio ("buy it direct in AI Studio").
Currently, official details on custom domains are limited to these two X posts, as the November 17 Google Developers Blog post focuses solely on Android and Google Play workflows. Pricing, renewal terms, bring-your-own-domain support, and eligible top-level domains have not been detailed.
Full-Stack Android Apps and Six New Features at a Glance
In the Google Developers Blog post authored by Mike Taylor-Cai, Google explains that AI Studio previously generated Android apps restricted to local session storage. Now, when a prompt calls for persistence, user sign-in, or generative AI, the Build agent provisions backend infrastructure upon user confirmation and wires it directly into production-ready Kotlin code. Google summarizes six core additions:
**Full-stack persistence and sign-in: **Automatically sets up Cloud Firestore (including Security Rules) and Google Sign-In.
**Secure Gemini API calls: **Routes Gemini requests through Firebase AI Logic and Firebase App Check (backed by Google Fraud Defense, formerly reCAPTCHA Enterprise) rather than embedding API keys inside the APK.
Personal Google Play publishing: **Ships Android apps directly to a user's **personal Google Play library using a standard Google account for one-click installation on their phone. Developers can set up a Play account from the publish panel while Google AI Studio builds and signs the Android App Bundle (.aab) behind the scenes (this targets personal device installation via a personal Play library rather than public Play Store listing).
**Faster, high-definition emulator with audio: **Delivers lower latency, an HD toggle up to 1200p, and real-time audio inside the browser.
**One-click screenshot to agent: **Lets developers send emulator screenshots directly to the Build agent with one click.
**Custom fonts and Android CLI integration: **Supports custom typography and integrates with the official Android CLI for project scaffolding and Gradle builds.

Image source: Google Developers Blog (build-full-stack-android-apps-with-gemini-in-google-ai-studio).
Regional Availability and Cost Caveats
On geographic availability, Logan Kilpatrick's X post explicitly notes that building an Android app and downloading it to a phone through the Google Play store is US only for now, whereas the Google Developers Blog post makes no mention of regional restrictions; both official statements should be kept in mind. On pricing, the blog states that developers can build and test on their own devices directly from the Play Store "at no cost," without publishing a broader commercial rate card for Firebase or Play services.
