AI
Kora aims to build an OS designed from the ground up for AI developers and power users
A new Linux-based desktop operating system launching in 2027 prioritizes seamless integration of multiple AI tools rather than mass-market appeal.
WIRED — AI
Kirt McMaster's upcoming Kora operating system takes a deliberately narrow approach: it's built for AI professionals and advanced users, not mainstream consumers. The Linux-based desktop OS, announced this week, aims to solve a core friction point in how people currently interact with AI tools—fragmenting them across browser tabs and isolated applications.
The AI-First Architecture#
Kora's central premise is that context matters more than compartmentalization. Rather than treating Claude, ChatGPT, and local models as separate desktop applications, the operating system integrates them at the system level. This means a designer could shift between an OpenAI model to brainstorm, switch to Anthropic's tools to write and deploy code, and tap into a locally-run model to cut token costs—all within a unified experience where the OS remembers and surfaces relevant context.
The "neural file system" extends this philosophy to how you find information. Instead of hunting for specific filenames, you can query your machine conversationally: ask for "that business contract with the late payment clause" and the OS indexes files by semantic meaning rather than metadata alone.
Hardware and Compatibility#
Kora will land on Snapdragon-powered PCs through a Qualcomm partnership starting next year, but the team is also targeting Intel and AMD-based machines. To bridge the software gap, built-in compatibility layers let users run existing Windows applications, though the startup is developing native apps for email, contacts, and calendaring. Developers will have access to an SDK for creating Kora-native software.
The system inherits gaming capabilities from McMaster's earlier venture, Playtron—its GameOS architecture (now called GameFrame within Kora) supports PC gaming on compatible hardware.
The Team's Pedigree#
McMaster brings experience from cofounding Cyanogen, the Android fork that attracted millions of users before shutting down roughly a decade ago. Cofounder Franck De Girolami worked on GameOS development, while John Lagerling contributed expertise from his time on Android at Google. The lean team—under 100 people—is betting they can move faster than incumbents locked into supporting billions of users.
McMaster argues that macOS and Windows face structural barriers to rethinking the desktop for AI-native workflows. Kora is optimizing for machines with at least 64GB of RAM, double the memory ceiling on Google's recently launched Googlebook, underscoring ambitions to run compute-heavy models locally.

