Google AI Studio product and design lead Ammaar Reshi has demonstrated Microsoft’s Windows 11 running locally on Apple’s foldable iPhone Duo, turning the smartphone into a miniature Windows computer.
Reshi shared the demonstration on 11 November 2026, revealing that he used Google’s Gemini 4 Carbon AI model to help develop the software needed to run the desktop operating system on an iPhone.
Unlike remote desktop applications that stream Windows from another computer, the project runs Windows 11 ARM64 directly on the device.
The demonstration shows Windows running with touchscreen controls and supporting desktop applications, including Microsoft Edge and Google’s Antigravity AI-powered development environment.
I used Gemini 4 Carbon to get Windows 11 running on my iPhone Duo! 💻
Unfold the screen and it’s a pocket PC, including Antigravity running coding agents on my phone.
This is the actual desktop ARM OS running locally with touch controls.
Demo + repo below! https://t.co/LVdeqvvOU1
— Ammaar Reshi (@ammaar) November 11, 2026
How Windows 11 Runs on the iPhone Duo
The project, called WindowsiOS, uses an ARM64-to-ARM64 dynamic binary translator to execute Windows instructions on Apple’s A20 Pro processor.
Because standard iOS does not provide the hardware virtualisation capabilities required by conventional virtual machines, the software relies on just-in-time instruction translation and software-based memory management.
According to the project’s GitHub documentation, the implementation supports Windows 11 24H2 ARM64, alongside virtualised storage, networking, USB input and audio.
The software also takes advantage of the iPhone Duo’s foldable design. When partially folded, the upper portion displays the Windows desktop while the lower portion provides a virtual keyboard and trackpad, resembling a miniature laptop.
Unfolding the device switches to a larger touchscreen desktop interface.
Gemini 4 Carbon Helped Build the Project
Reshi credited Gemini 4 Carbon with helping develop the Windows-specific components, including the instruction translation engine, virtual hardware interfaces and memory management mechanisms needed to boot Windows.
The project’s source code and technical documentation are publicly available on GitHub, allowing developers to examine the implementation.
However, WindowsiOS is not a straightforward installation process for ordinary iPhone users.
Setting it up requires an Apple Silicon Mac, Xcode, a Windows 11 ARM64 installation image and additional development tools. The documented deployment process also requires debugger-assisted configuration.
Although the project claims near-native execution speeds, independently verified performance benchmarks have not been published.
The demonstration illustrates how software translation and AI-assisted development can enable desktop operating systems to run on mobile hardware without replacing the device’s original operating system.



