OpenAI Ships ChatGPT Desktop App to Linux with Codex in Preview
OpenAI launched the ChatGPT desktop app for Linux in preview, bundling ChatGPT, ChatGPT Work, and Codex into one native experience across Ubuntu, Debian, and Fedora.
On August 11, 2026, OpenAI quietly shipped something the Linux community had been requesting for years: a native ChatGPT desktop app. The preview release, announced via OpenAI’s official X account, brings the full ChatGPT, ChatGPT Work, and Codex experience to Linux as a single unified application. This marks the first time OpenAI has offered an official desktop client for the platform that powers the vast majority of AI development infrastructure worldwide.
What’s in the Box
The ChatGPT desktop app for Linux isn’t a stripped-down port. It’s the same workspace-oriented application that macOS and Windows users have been running, now compiled and tested for the Linux ecosystem. The app combines three distinct experiences into one native binary:
ChatGPT — the conversational AI assistant for general queries, writing, analysis, and research. The Linux version includes the in-app browser and Chrome extension support that debuted on other platforms, meaning users can interact with web content directly within the app’s workflow pipeline.
ChatGPT Work — the productivity-focused variant designed for professional use cases, with project management, long-running work sessions, and file handling capabilities. The desktop app is explicitly designed as a “workspace for managing projects, working with files, and using browser workflows.”
Codex — OpenAI’s AI coding agent, now natively integrated into the desktop experience on Linux. Codex can run alongside ChatGPT, analyze codebases, propose changes, and execute development tasks. This is arguably the most significant part of the launch for the Linux developer community, which has historically been underserved when it comes to first-party AI coding tools.
Supported Distributions and Architectures
OpenAI has been deliberate about which Linux distributions to support at launch. The preview targets the most widely used and stable enterprise and community distributions:
- Ubuntu 24.04 LTS and Ubuntu 26.04 LTS — the backbone of enterprise and cloud Linux
- Debian 13 — the stable upstream for countless downstream distributions
- Fedora 43 and Fedora 44 — the cutting-edge community distribution
Both x64 and ARM64 architectures are supported, a critical detail given the growing prevalence of ARM-based development hardware and the dominance of ARM in the single-board computer and edge computing spaces. Packages are distributed in .deb and .rpm formats, covering the two major packaging ecosystems.
Why This Matters
The significance of this launch extends well beyond the convenience of having an official app. Linux is the operating system that runs the vast majority of AI training and inference infrastructure. The servers that train models, the workstations that researchers use, the cloud instances that developers provision — they overwhelmingly run Linux. Until now, OpenAI’s developer-facing tools on Linux were limited to the Codex CLI, which requires a terminal and command-line proficiency. The desktop app brings a full GUI experience with integrated workflows that were previously only available on macOS and Windows.
The inclusion of Codex in the desktop app is particularly notable. Codex is OpenAI’s AI coding agent, capable of understanding entire codebases, proposing multi-file changes, and executing complex development tasks. Having it natively available on Linux — the primary platform for software development — means that developers can now use Codex in the same environment where their code actually runs. This eliminates the friction of context-switching between a macOS or Windows machine for AI-assisted development and a Linux server for deployment and testing.
The browser workflow integration is another key feature. The desktop app includes an embedded browser that supports Chrome extensions, allowing users to chain together research, analysis, and action across web content. On Linux, where browser-based developer tooling is ubiquitous, this creates a seamless loop between web research and AI-powered execution.
The Unofficial Precedent
Before this official release, the Linux community had already been finding workarounds. Several unofficial projects appeared on GitHub, including codex-desktop-linux wrappers that converted OpenAI’s upstream macOS DMG into native Linux Electron apps with .deb, .rpm, AppImage, and Nix packaging. These community efforts demonstrated clear demand and provided a testing ground for what an official Linux release would need to address. OpenAI’s official app supersedes these unofficial wrappers with proper packaging, official support channels, and the full feature set — including ChatGPT Work, which was not available in most community ports.
The Bigger Picture
This launch is part of a broader strategy by OpenAI to make its tools available everywhere developers work. The company has been systematically expanding platform support: Codex launched on macOS, then Windows in early 2026, and now Linux. The ChatGPT desktop app follows the same trajectory. By covering all three major desktop operating systems, OpenAI ensures that its AI tools are available regardless of a developer’s platform preference.
For the Linux ecosystem specifically, this is a meaningful signal of vendor commitment. Historically, many AI companies have treated Linux as a second-class platform, offering CLI-only tools or web interfaces while reserving full-featured desktop applications for macOS and Windows. OpenAI’s decision to invest in a native Linux desktop app — with both x64 and ARM64 support, multiple distribution targets, and both major package formats — suggests a recognition that Linux developers are a core audience worth investing in directly.
The timing is also strategic. With AI coding tools becoming increasingly competitive — Anthropic’s Claude Code, Google’s Gemini Code Assist, and open-source alternatives like Continue all vying for developer mindshare — having a polished, native desktop experience on Linux gives OpenAI a meaningful distribution advantage among the developers who are most likely to influence enterprise tooling decisions.
What’s Next
The app is currently in preview, which means users should expect rough edges, potential stability issues, and a feedback-driven improvement cycle. OpenAI has opened a thread on its developer community forum specifically for the Linux preview, indicating an active interest in collecting user feedback. As the preview matures, it’s reasonable to expect expanded distribution support, additional package formats, and tighter integration with Linux-specific developer workflows.
For now, Linux developers can download the app directly from chatgpt.com/codex and start using ChatGPT, ChatGPT Work, and Codex in a unified native desktop experience — something that, until this week, simply didn’t exist.
Sources
- [1] https://x.com/OpenAI/status/2087231350134980830
- [2] https://www.phoronix.com/news/ChatGPT-Desktop-Linux-Preview
- [3] https://thenewstack.io/openais-chatgpt-desktop-linux/
- [4] https://community.openai.com/t/codex-in-chatgpt-desktop-app-for-linux-is-now-in-preview/1390027
- [5] https://www.thurrott.com/a-i/openai-a-i/340330/openai-brings-chatgpt-and-codex-to-linux
- [6] https://aiweekly.co/alerts/openai-ships-chatgpt-desktop-app-to-linux-with-codex-in-preview
- [7] https://learn.chatgpt.com/docs/app