The merger of ChromeOS and Android has been reported, denied, half-confirmed and reported again since Android Authority first described the project. At Qualcomm’s Snapdragon Summit, two things changed. Google named the platform — Android for PC — and it attached a year to it.
What was actually said
Rick Osterloh, Google’s head of platforms and devices, framed the shift as the end of a split the company has maintained for fifteen years: separate systems for phones and for computers, replaced by what he called a common technical foundation serving both. He went further than the architecture, committing to bring Gemini and Google’s full Android AI stack, the Play Store catalogue and the Android developer community into the PC operating system. Sameer Samat, who runs the Android ecosystem, supplied the timing: next year, which means 2026.
Cristiano Amon, Qualcomm’s chief executive, said he had seen it running and called it incredible — a CEO on his own stage, so discount accordingly, but a working build seen by a partner is still a working build. The business side of the announcement is blunter than the technical side. Chromebooks today are overwhelmingly Intel machines. Android for PC is being built on Snapdragon, on the back of the Snapdragon X line Qualcomm has spent two years pushing into Windows laptops with modest success. Google is changing which chip vendor its laptop platform belongs to.
Why this attempt is different
Google has tried this before. Android apps came to Chrome OS in a container in 2016 and never felt native. Andromeda, an explicit Android–Chrome OS merger, was shelved the same year. Fuchsia, a from-scratch operating system, shipped quietly on smart displays and went no further. The reasonable instinct is to treat the fourth attempt like the first three.
What is different is that most of the work has already shipped, in Android itself, for other reasons. Android now has real desktop windowing — resizable, overlapping windows with a taskbar — developed for foldables and for phones driving external monitors. It handles external displays, keyboards, trackpads and mice as first-class inputs because tablets forced it to. In 2025 it gained a Linux terminal running a virtual machine, which is the single feature that separates a browsing appliance from a developer’s machine. None of that was built for laptops, and all of it is what a laptop needs. The merger is less a new operating system than the point at which Android stops pretending those features are for something else.
The three problems
Applications. A Play Store app is built for a touchscreen held in one hand. Phone apps running full-screen on a 14-inch display with a keyboard is the exact failure that made Android tablets a joke for a decade. Google’s answer is adaptive layouts and developer requirements, and its leverage over developers is real — but the catalogue it inherits on day one was not written for this.
Existing Chromebooks. There are tens of millions in schools, most of them Intel, many of them already past their automatic-update expiry. Google has promised nothing about whether they get Android for PC, a long ChromeOS support tail, or a quiet end. For the institutions that bought them, that is the only question that matters.
The reason to switch. ChromeOS failed outside American and European schools, and cheap Windows laptops won the rest of the market. For a buyer deciding between a Tk 35,000 Windows machine and a Tk 35,000 Android one, app compatibility with the phone they already own is a genuine argument; so is not paying for a Windows licence. Whether it beats the thing everyone’s workplace already runs on is a different question, and Google has not answered it yet.
What to watch
Two signals will tell you whether 2026 is real. The first is a developer preview with the desktop form factor named in it, rather than a tablet build with windowing switched on. The second is an OEM other than Google committing to a Snapdragon Android laptop with a price — Lenovo, Acer and Asus all build both Chromebooks and Windows machines, and they will not announce a third line without volume forecasts. Until one of them does, the announcement is a direction, not a product.
