Google AI Studio sells domains from $12 a year and puts Android apps on your phone
In Build mode, where AI builds a website or app from a description, Google AI Studio now sells domains itself and installs Android apps on a phone. Previously, a site lived at a free *.ai.studio address, and you had to buy a custom domain separately and configure it by hand. Now the Publish panel offers domains from ~$12 a year, and the service sets up DNS and the SSL certificate on its own. For now this works only in the US and requires a paid Google Cloud project.
AI-processed from TG: @cgevent (Метаверсище и ИИще); edited by Hamidun News
Google AI Studio's Build mode, where AI builds a website or mobile app from a text description, has gained two features for moving from prototype to real product. First, you can now buy a domain for your site directly in the publishing panel, from ~$12 a year depending on the top-level domain. Previously, a project sat at a free address like *.
ai.studio, and you had to register your own name with a third-party registrar and connect it to the server yourself. Second, a finished Android app can be loaded onto a real smartphone, rather than only viewed in an emulator.
All of this is available only in the US for now, according to the author of the Telegram channel «Метаверсище и ИИще», who reported the news.
A domain without the configuration hassle
Imagine you have put together a service MVP in one evening and want to show it to clients at your brand's address rather than at a *.ai.studio link. Now, in the Publish panel, you just click "+ Add custom domain", select "Buy a new domain" and pay for the name without leaving the interface.
The service handles the rest. It registers the name in Cloud Domains, Google's domain registrar. Then it sets up a route in Cloud DNS so that requests to the domain reach the Cloud Run backend where your app runs. After that, it issues a managed SSL certificate, which lets the site open over secure HTTPS. Before, this was a chain of steps across a registrar, a DNS panel and a wait for the certificate to be verified. The old *.ai.studio address keeps working, so links you have already shared won't break.
The requirements: a paid Google Cloud project with active Cloud Billing and Owner or Editor permissions. The feature is not available on the free Starter Tier. For a student who just wants to try out the service, that is a barrier. For a team with a working cloud account, almost nothing changes.
An Android app on a real phone
The mobile branch of Build generates native Android apps in Kotlin and Jetpack Compose. The backend is built on Firebase: Cloud Firestore stores data, Google Sign-In handles login, and App Check verifies that requests come from the genuine app. AI features run through Firebase AI Logic, and you don't need to hard-code API keys into the APK, the app's installation file. This matters because outsiders can extract a key embedded in an app.
Until now, you could only view the result in the built-in browser emulator. Now you can load the generated app onto a physical smartphone. The post's author still does this the old way, using ADB (the Android debugging tool), developer mode and a USB cable. For testing, this is significant: the camera, gestures, battery drain and response speed can behave differently on real hardware than in an emulator.
What this means in practice
For a founder or product manager, this is a step toward taking an idea from prompt to public link in a single environment. You can show the site on your own domain and test the mobile version on your own phone.
The author argues that Google is using the advantage of its ecosystem: the company already has domains, DNS, hosting, authentication and payments, and it is connecting them to its AI products. He compares this to the wallet OpenAI announced: by his account, Google's reply was that it has had its own wallet for about fifteen years. That is the author's opinion, but the trend is visible: competition is shifting from model quality to who has built the better set of services around the model.
It's worth considering the limitations up front:
- the features are available only in the US for now, according to the author;
- you need a paid Google Cloud project with billing enabled;
- the project is tied to Cloud Run, Firebase and other Google services, and migrating later will take work;
- the post says nothing about installing on a phone without a cable and ADB.
If you already work in Google Cloud and are in a supported region, try buying a test domain and see how quickly HTTPS comes up. For everyone else, it's enough to know that the path from "generated an app" to "showed it to people" in AI Studio has gotten shorter.
- Nov 19, 2026RAG Knowledge Base: 100 GB of Files Yield 1.5 Million Chunks and 9 Hours of Loading
- Nov 19, 2026AI Models Added a School to the Target List in 70% of Vals AI Tests
- Nov 19, 2026Q.ANT and AMD Run a 1B-Parameter LLM With an Optical Chip in Place of Attention
- Nov 19, 2026The Word "Video Game" Raised GPT-6 Astra's Willingness to Pick Targets From 6% to 92%
Key Questions Answered
What are the key details on A domain without the configuration hassle?
Imagine you have put together a service MVP in one evening and want to show it to clients at your brand's address rather than at a .ai.studio link. Now, in the Publish panel, you just click "+ Add custom domain", select "Buy a new domain" and pay for the name without leaving the interface. The service handles the rest. It registers the name in Cloud Domains, Google's domain registrar. Then it sets up a route in Cloud DNS so that requests to the domain reach the Cloud Run backend where your app runs.
What are the key details on An Android app on a real phone?
The mobile branch of Build generates native Android apps in Kotlin and Jetpack Compose. The backend is built on Firebase: Cloud Firestore stores data, Google Sign-In handles login, and App Check verifies that requests come from the genuine app. AI features run through Firebase AI Logic, and you don't need to hard-code API keys into the APK, the app's installation file. This matters because outsiders can extract a key embedded in an app. Until now, you could only view the result in the built-in browser emulator.
What this means in practice?
For a founder or product manager, this is a step toward taking an idea from prompt to public link in a single environment. You can show the site on your own domain and test the mobile version on your own phone. The author argues that Google is using the advantage of its ecosystem: the company already has domains, DNS, hosting, authentication and payments, and it is connecting them to its AI products.
Test Frontier LLMs & Autonomous Agents in One Workspace
Run Claude Opus 5.5, GPT-6, Gemini 3.1, and DeepSeek side-by-side in Nomad AI (nomadnet.ai) — unified model routing, deep research, and multimodal workflows under one subscription.
The AI world, distilled — once a week
Seven stories that actually mattered, hand-picked. No noise, no reposts, no press releases.
Done! Check your inbox for a confirmation.