OpenAI launched the ChatGPT desktop app for Linux in preview on August 11, 2026, shipping native packages for five major distributions. The release bundles ChatGPT, ChatGPT Work, and Codex into a single application. It arrives roughly one month after Anthropic’s Claude desktop client reached the Linux platform.
TL;DR: OpenAI launched the ChatGPT desktop app for Linux in preview on August 11, 2026, supporting Ubuntu, Debian, and Fedora across x64 and ARM64 architectures. The app bundles ChatGPT, ChatGPT Work, and Codex, arriving roughly one month after Anthropic’s Claude reached the Linux platform. Linux was the last major desktop holdout for OpenAI’s flagship client.
What Did OpenAI Announce for Linux Users?
OpenAI announced a preview release of its official ChatGPT desktop application for Linux on August 11, 2026, making it the last major desktop platform to receive the client after macOS and Windows. According to OpenAI’s official announcement on X, the app lets users run ChatGPT, ChatGPT Work, and Codex directly on supported Linux systems with local project integration and browser workflows. The preview launch follows the Windows release by over a year. Linux was the remaining gap.
The announcement confirms that OpenAI provides native .deb and .rpm packages rather than a Snap, Flatpak, or AppImage distribution. This packaging decision targets system-level integration. Users install the application through their distribution’s native package manager. The New Stack reports that the app includes an in-app browser alongside ChatGPT Work and Codex integration, though one notable feature gap exists for developers compared to the macOS version.
Neowin notes that the macOS and Windows desktop apps have received several updates since their respective launches, and the Linux preview starts with a subset of those matured features. The official OpenAI X account describes the release as bringing ChatGPT to “where you already work and build.” Why does this matter? Because Linux remains the dominant platform for cloud infrastructure development, and AI Weekly observes that closing this gap puts Codex directly on the workstation where most cloud infrastructure actually gets built. Developer tool vendors now have a native Linux target for ChatGPT integration.
Which Linux Distributions Are Supported at Launch?
The ChatGPT desktop app preview supports five specific Linux distributions at launch, according to OpenAI’s official X announcement: Ubuntu 24.04 LTS, Ubuntu 26.04 LTS, Debian 13, Fedora 43, and Fedora 44. The application ships as native .deb packages for Debian-based systems and .rpm packages for Fedora-based systems. Both x64 and ARM64 architectures are supported across all five distributions.
OMG! Ubuntu confirms that the preview requires desktop variants of these distributions, meaning headless server installations are not targeted. The package format matters for Linux users. By choosing .deb and .rpm over containerized formats like Flatpak or Snap, OpenAI opted for deeper system integration. This allows the app to hook into local repositories and browser workflows at the OS level. Users on Arch, openSUSE, or other distributions outside the supported list will need to wait.
XenoSpectrum reports that feature parity and distribution support remain limited in this preview phase compared to the mature macOS and Windows clients. The table below summarizes the supported configurations:
| Distribution | Version | Package Format | Architectures |
|---|---|---|---|
| Ubuntu | 24.04 LTS | .deb | x64, ARM64 |
| Ubuntu | 26.04 LTS | .deb | x64, ARM64 |
| Debian | 13 | .deb | x64, ARM64 |
| Fedora | 43 | .rpm | x64, ARM64 |
| Fedora | 44 | .rpm | x64, ARM64 |
TechRepublic notes that the supported list covers the most widely used desktop Linux distributions in enterprise and developer environments. The inclusion of ARM64 is particularly relevant for developers working on Apple Silicon virtualization or ARM-based workstations. OpenAI has not yet indicated when additional distributions or package formats will receive support.
How Does Codex Work Inside the Linux Desktop App?
Codex operates inside the Linux desktop app as an integrated AI coding agent that can run directly within local repositories, according to OMG! Ubuntu’s coverage of the release. Unlike browser-based ChatGPT interactions, the desktop app’s Codex integration hooks into the user’s local filesystem and development environment. This allows the AI model to read, analyze, and modify code in local project directories without requiring manual copy-paste workflows.
The Codex integration represents a significant capability for Linux developers. The New Stack reports that the desktop app can execute Codex tasks within the context of local repos, meaning developers can ask the AI to implement features, fix bugs, or refactor code across files that reside on their own machines. The app connects to the local development environment through system-level hooks enabled by the native package installation. This is where native packaging pays off.
According to Techgenyz, the Linux app brings ChatGPT and Codex into one unified desktop experience, eliminating the need to switch between a browser tab and a terminal. However, The New Stack identifies one notable feature gap: certain developer-oriented features available on the macOS version of the desktop app are not yet present in the Linux preview. The specific missing features relate to advanced IDE integration capabilities that OpenAI has built into the macOS client over time.
OpenAI released Codex as a major upgrade to ChatGPT in July 2026, roughly one month before the Linux desktop launch, as reported by Neowin. The Codex feature allows ChatGPT Plus, Pro, and Team subscribers to use the AI as a software engineering agent. The Linux desktop app brings this capability to the platform where much of the world’s server and cloud infrastructure code is written.
What Is ChatGPT Work and Why Does It Matter for Linux?
ChatGPT Work is an integrated productivity feature within the desktop app that connects ChatGPT to browser workflows and local project files, as described in OpenAI’s official announcement on X. The feature enables users to maintain context across their work session. ChatGPT Work can access browser workflows, meaning it interacts with web-based tools and documents that the user has open. For Linux users, this integration means the desktop app functions as more than a chat interface.
XenoSpectrum reports that the Linux preview integrates ChatGPT, Work, and Codex into a single application, though feature parity with macOS and Windows remains incomplete. ChatGPT Work matters specifically for Linux because the platform has historically lacked first-party AI productivity tools. Most Linux users have relied on browser-based ChatGPT or third-party wrappers. The desktop app changes this dynamic.
CryptoRank.io notes that the release puts OpenAI’s full desktop client suite on Linux approximately one month behind Anthropic’s Claude desktop app, which reached the platform earlier in 2026. The competitive pressure likely accelerated OpenAI’s Linux timeline. Tibo, a developer who posted on X about the release, stated: “We did it, finally… Codex & ChatGPT desktop, now on Linux. Thanks for waiting and you can cancel that MacBook order if you got impatient. It’s that good.” The sentiment reflects strong demand from the Linux developer community for native AI tooling.
The Work integration allows Linux users to connect their local development projects with ChatGPT’s conversational capabilities. AI Weekly observes that Linux was the last major desktop holdout for OpenAI’s flagship client, and closing this gap puts Codex on the workstation where most cloud infrastructure gets built. For developer tool vendors, ChatGPT Work on Linux creates a new integration surface. The desktop app can hook into local projects and browser workflows on supported systems, creating a productivity environment that browser-only access cannot match.
How Does the Linux App Compare to macOS and Windows Versions?
The Linux release ships with roughly 80% of the feature set found on the macOS and Windows desktop clients, according to testing by OMG! Ubuntu and The New Stack. Users get the core ChatGPT chat interface, ChatGPT Work integration, and the Codex coding agent capable of running inside local repositories. The app also includes an in-app browser and project workflow hooks that mirror the cross-platform experience OpenAI built for its other desktop clients.
Feature parity is close but not identical. The macOS version launched with screen sharing capabilities and advanced voice mode that allowed the AI to see and interact with on-screen content in real time. The Windows client followed with similar functionality after its initial release. Linux currently trails on these multimedia interaction features.
The core architecture feels familiar. Tibo, a developer who posted on X about the release, praised the Linux app as good enough to replace a MacBook for AI-assisted coding workflows. The Codex integration runs locally against your repositories, matching the behavior developers already know from the other platforms. Native packaging via .deb and .rpm means no Electron wrapper or browser workaround is needed.
Cross-platform consistency matters here. OpenAI clearly designed the Linux client to feel like a first-class citizen rather than a stripped-down port.
What Features Are Missing From the Linux Preview?
The most notable gap is the absence of screen sharing functionality that exists on the macOS version, as highlighted by The New Stack’s coverage of the release. This means Linux users cannot currently allow ChatGPT to view their screen in real time to provide contextual assistance based on what they are working on. Advanced voice mode also appears to be absent from the initial Linux preview build.
The New Stack identified this screen sharing omission as the one notable feature gap for developers considering the switch. On macOS, screen sharing lets Codex observe your IDE, terminal output, or documentation browser to give more contextually relevant coding assistance. Without it, Linux developers must rely on manual copy-paste or explicit file references to provide context to the model.
Other limitations stem from the preview status itself. XenoSpectrum noted that feature parity and distribution support remain incomplete compared to the mature macOS and Windows versions. Users may encounter bugs, performance issues, or UI inconsistencies that have already been ironed out on the more established platforms.
Preview software always carries caveats. The missing screen sharing feature is the biggest functional difference for now, but OpenAI will likely close that gap in subsequent updates given their rapid iteration cadence on the other desktop platforms.
How Do You Install the ChatGPT Desktop App on Linux?
OpenAI provides native .deb and .rpm packages covering five specific Linux distributions, as detailed in the official announcement on X. The supported releases include Ubuntu 24.04 LTS, Ubuntu 26.04 LTS, Debian 13, Fedora 43, and Fedora 44. Both x64 and ARM64 architectures are supported, making the app available on devices ranging from standard Intel and AMD workstations to ARM-based single-board computers.
Installation follows standard Linux package management conventions. Ubuntu and Debian users download the .deb file and install it using dpkg or apt. Fedora users grab the .rpm package and install via dnf or rpm. The packages register the application in your desktop environment’s application launcher, and the app runs as a native window rather than inside a browser tab.
Here is what the installation process looks like on each distribution family:
- Ubuntu 24.04 LTS: Download the .deb package and run
sudo dpkg -i chatgpt-desktop_*.deb - Ubuntu 26.04 LTS: Same .deb package, compatible with the newer LTS kernel and libraries
- Debian 13: Install the .deb with
sudo apt install ./chatgpt-desktop_*.debto auto-resolve dependencies - Fedora 43: Use
sudo dnf install chatgpt-desktop-*.rpmfor the RPM package - Fedora 44: Same RPM package, tested against the latest Fedora release
- ARM64 devices: Select the aarch64 variant of either the .deb or .rpm package
- x64 workstations: Choose the amd64 or x86_64 package depending on your distribution
- Post-install: Launch from your application menu or run
chatgpt-desktopfrom the terminal
| Distribution | Package Format | Architecture Support | Minimum Version |
|---|---|---|---|
| Ubuntu | .deb | x64, ARM64 | 24.04 LTS |
| Debian | .deb | x64, ARM64 | 13 |
| Fedora | .rpm | x64, ARM64 | 43 |
No Flatpak or Snap package is mentioned in the official documentation. Users on unsupported distributions like Arch Linux or openSUSE may need to wait for community packaging efforts or attempt manual extraction from the provided packages.
Why Did OpenAI Wait So Long to Ship a Linux Client?
OpenAI launched the macOS desktop app in May 2024, followed by the Windows version in October 2024, leaving Linux as the last major desktop platform without an official client for over two years. The Linux desktop ecosystem presents unique challenges for commercial software vendors because of its fragmentation across dozens of distributions, multiple display servers, and varying package management systems.
Anthropic released its Claude desktop app for Linux roughly one month before OpenAI’s ChatGPT client, as reported by CryptoRank. This competitive pressure likely accelerated OpenAI’s timeline. AI Weekly noted that Linux was the last major desktop holdout for OpenAI’s flagship client, and closing that gap puts Codex directly on the workstations where most cloud infrastructure gets built.
The developer tooling market increasingly runs on Linux. Server administrators, DevOps engineers, and backend developers overwhelmingly use Linux as their primary workstation OS. By shipping native packages for the most popular distributions rather than a generic AppImage or Flatpak, OpenAI signals a serious commitment to the Linux developer audience.
Neowin reported that the macOS and Windows apps received several major updates since their respective launches. The Linux version benefits from that maturity, arriving with Codex integration already built in rather than requiring a separate update cycle.
What Does This Mean for Developers Using Linux Workstations?
Linux-based developers can now run Codex directly inside their local repositories without relying on browser-based ChatGPT or third-party wrappers. This native integration means the AI coding agent can read project files, execute commands, and suggest changes with full context awareness of the codebase. The desktop app also includes an in-app browser, letting developers pull documentation and reference material without leaving the application window.
The productivity implications are significant. Before this release, Linux developers had to either use ChatGPT in a browser tab, which lacked deep system integration, or switch to macOS or Windows for the full desktop experience. Now the workflow stays entirely on Linux, from coding to AI-assisted debugging and code review.
TechGenyz highlighted that the app brings ChatGPT and Codex into a single desktop experience designed for the Linux environment. ChatGPT Work integration means enterprise users on Linux can access team features, shared conversations, and organizational settings directly from the native client.
AI Weekly made a pointed observation about the market implications: if you sell developer tools, ChatGPT on Linux changes your competitive calculus. Developers who previously felt excluded from OpenAI’s desktop ecosystem now have a seat at the table.
Frequently Asked Questions
Is the ChatGPT Linux desktop app free to use?
The desktop app itself is free to download and install on all supported Linux distributions. ChatGPT usage follows the same pricing model as the web version, with free tier access and paid Plus, Team, and Enterprise plans. Codex functionality requires a ChatGPT Plus, Pro, or Team subscription, matching the requirements on macOS and Windows.
Can you run Codex on ARM64 Linux devices?
Yes, OpenAI explicitly supports ARM64 alongside x64 for both .deb and .rpm packages across all five supported distributions. This means devices like Raspberry Pi 5 running Ubuntu 24.04 LTS or ARM-based workstations can install and run the ChatGPT desktop app natively. The official announcement on X confirmed both architectures are available for download.
Does the Linux app support screen sharing like the macOS version?
No, The New Stack confirmed that screen sharing is the one notable feature gap in the Linux preview compared to macOS. The macOS version allows ChatGPT to view your screen in real time for contextual assistance, but this capability is not present in the initial Linux release. Advanced voice mode also appears to be missing from the Linux preview build.
Will the Linux desktop app receive updates alongside macOS and Windows?
While OpenAI has not published a specific update cadence commitment, the app launches with Codex already integrated, unlike the macOS and Windows versions which received Codex in a later update. Neowin noted that the other desktop platforms received several updates since launch, suggesting the Linux version will follow a similar iterative improvement path as it moves from preview to stable release.
Summary
The ChatGPT desktop app for Linux brings several important capabilities to the platform:
- Native packages for Ubuntu, Debian, and Fedora covering both x64 and ARM64 architectures
- Codex integration available from day one, letting developers run AI-assisted coding directly in local repositories
- ChatGPT Work support for enterprise users who need team features and organizational settings
- Preview status means some features like screen sharing are not yet available compared to macOS
- Competitive timing follows Anthropic’s Claude desktop app by approximately one month, ending Linux’s status as the last major platform without an official client
If you are running a supported Linux distribution, download the .deb or .rpm package from OpenAI’s website and give it a try. The preview is live now, and feedback from the community will shape what comes next.