5 Key Takeaways
- OpenAI launched a ChatGPT desktop app for Linux in preview, extending its desktop tools to all major operating systems.
- The Linux version lacks native Computer Use, so it cannot control desktop applications like OpenOffice or GIMP at launch.
- The app supports core ChatGPT, Codex, Voice, an in-app browser, and Chrome extensions on Ubuntu, Debian, and Fedora via .deb and .rpm packages.
- The desktop app complements existing Linux options like the Codex CLI and IDE extensions by adding a visual workspace for coordinating tasks.
- OpenAI has not announced a timeline for bringing Computer Use to Linux, but browser-based workflows are supported and the preview is expected to evolve.
OpenAI’s ChatGPT Desktop App Lands on Linux—but Not Every Feature Made the Trip
OpenAI has brought its ChatGPT desktop application to Linux. The app, now in preview, combines ChatGPT, ChatGPT Work, and Codex into a single interface. With this release, OpenAI says its desktop tools now span every major desktop operating system. But the Linux version arrives with a notable limitation: it cannot yet control desktop applications outside its built-in browser.
The macOS and Windows versions of the ChatGPT desktop app launched in July, tied to the debut of ChatGPT Work, OpenAI’s product aimed at workplace and knowledge tasks. Since then, Linux users have repeatedly asked for a native version. That request has now been answered, at least in part.
A Long-Requested Release
The Linux preview supports Ubuntu 24.04 and 26.04 LTS, Debian 13, and Fedora 43 and 44. The term LTS stands for Long-Term Support, meaning those Ubuntu releases receive maintenance updates for an extended period. OpenAI is providing native installation packages in .deb and .rpm formats for both x64 and ARM64 architectures.
For readers less familiar with Linux software distribution, .deb files serve Debian-based systems such as Ubuntu, while .rpm files serve Fedora and related distributions. The x64 build covers traditional Intel and AMD processors, while ARM64 covers newer ARM-based processors. This packaging approach should make installation straightforward on most mainstream desktop Linux setups.
An OpenAI spokesperson said:
“Linux has been one of the most-requested platforms for the desktop app, and this launch extends ChatGPT and Codex across every major desktop operating system.”
That is a meaningful milestone. Developers and enterprise users who work primarily on Linux no longer have to rely only on a web browser or command-line tools to use OpenAI’s desktop experience. The app will be available globally at launch.
What’s Included—and What’s Missing
Like the ChatGPT desktop app on macOS and Windows, the Linux version includes OpenAI’s in-app browser and support for Chrome extensions. It also supports the core ChatGPT and Codex experience, including voice.
The main gap involves a capability called Computer Use. In simple terms, Computer Use allows ChatGPT to interact with programs outside the browser—such as clicking, typing, and navigating inside desktop applications like OpenOffice or the image editor GIMP. That feature is not available on Linux at launch.
The same spokesperson explained the difference in detail:
“The Linux app supports the core ChatGPT and Codex experience, including Voice. The main difference is that native Computer Use — the ability for ChatGPT to interact with desktop applications like OpenOffice or GIMP — isn’t available at launch. This also affects related features, including Appshots, Record & Replay, and voice requests involving desktop apps and browser. Other browser-based workflows, including the Chrome extension, are supported.”
So while browser-based automation works, the deeper system-level control that can operate desktop software directly is absent for now. Features that depend on that ability—such as Appshots, Record & Replay, and certain voice requests—are also affected.
From Chat Window to Project Dashboard
When OpenAI first launched the new ChatGPT desktop app, it felt somewhat like a rebranded Codex application with ChatGPT Work and the standard ChatGPT chat attached. The company has since refined the experience, making it easier to switch between ChatGPT and Codex. Still, casual users may find the combined interface a little confusing.
OpenAI is not alone in facing this challenge. Anthropic’s Claude desktop app has similar growing pains. Across the industry, finding the best user experience for agentic systems—software that can plan and carry out multi-step tasks—remains a work in progress.
Chat windows are excellent when a user wants to ask a question and receive an answer. They become less natural when an agent runs for an hour, coordinates several subtasks, waits for approval, edits files, and produces multiple outputs. At that point, the product needs to behave as much like a project dashboard as a messaging app.
Where the Desktop App Fits
It is worth noting that Linux users have not exactly lacked access to OpenAI’s coding agent. The Codex CLI already runs on Linux. A CLI, or command-line interface, lets developers type commands directly into a terminal. OpenAI also offers an IDE extension, which connects Codex to code editors, and Codex cloud, which allows coding work to be delegated without installing a desktop client.
What the desktop app adds is a visual workspace for coordinating that work. Users can keep multiple projects and long-running tasks in view, inspect files and generated artifacts, review code changes, and move between local and cloud-based tasks. The app also brings OpenAI’s browser, plugins, and broader ChatGPT features into the same interface.
For developers who live in the terminal, the new app will not necessarily replace the CLI. The CLI remains the more direct option for inspecting a repository, running commands, scripting repeatable tasks, and integrating Codex into continuous integration—commonly called CI. The desktop app is aimed at the coordination layer around those tasks, especially when several agents, repositories, or different kinds of work are involved.
Because the app also includes ChatGPT Work for knowledge tasks and the standard ChatGPT chat, the Linux release is more than a coding tool. It gives users a single place to move between conversation, research, and agent-driven coding.
What Happens Next
The Linux launch means OpenAI now has a desktop presence on macOS, Windows, and Linux. That matters for developers who use Linux as their primary environment and for organizations that standardize on it.
The missing native Computer Use feature, however, may give some teams pause if they hoped to automate actions in desktop applications like LibreOffice or GIMP. OpenAI says other browser-based workflows, including the Chrome extension, are supported, so many web-based tasks should still work.
The company has not announced a timeline for bringing native Computer Use to Linux. Given that this is a preview, the app will likely continue to evolve. For now, Linux users get most of the ChatGPT desktop experience—just not the part that reaches outside the browser.
No comments:
Post a Comment