[Feature]: No Linux support is simply unreasonable.

[enihcam]

Problem / 你遇到的问题

No linux.

Proposed solution / 期望方案

Linux.

Platform / 平台

Linux desktop support

Additional context / 补充信息

I've searched existing issues, but the voice needs to be louder.

Before submitting / 提交前确认

• I have searched existing issues.

[reneleonhardt]

Linux is very diverse, could you specify which distribution support would be desired?
Ubuntu or Fedora?
KDE or Gnome?
X11 or Wayland?
Native packages or Flatpak?

[SuarezPM]

Linux is very diverse, could you specify which distribution support would be desired? Ubuntu or Fedora? KDE or Gnome? X11 or Wayland? Native packages or Flatpak?

CachyOS
KDE plasma 6
Wayland
Native packages

[enihcam]

Linux is very diverse, could you specify which distribution support would be desired? Ubuntu or Fedora? KDE or Gnome? X11 or Wayland? Native packages or Flatpak?

1. For me, Arch-based first, e.g. CachyOS.
2. It should be universal, so KDE/Gnome/LXQT/COSMIC/HYPRLAND etc. do not actually matter, but strongly prefer Qt6 or GTK3+ (please DON'T build upon Qt5 or Gtk4).
3. Don't worry. Let the communities do the packaging for you, and NO Flatpak please.

[abderaouflatr]

@reneleonhardt To answer your question on diversity, here is a concrete data point from a daily driver DevOps/IT Management setup.

Adding my strong support for this feature. The lack of a native Linux client severely breaks the workflow for developers and IT managers who rely on local homelab integrations, native terminal commands, and seamless file system access. Switching to a dual-boot (Windows) just to use the desktop client defeats the purpose of an integrated AI coding assistant.

Technical Environment Requirements

Here are the specific technical requirements for my environment to help guide the packaging:

• Distribution: Ubuntu 26.04 LTS (Debian-based)
• Display Server: Wayland
• Desktop Environment: GNOME (Ubuntu Default)
• Kernel: 7.0.0-22-generic
• Hardware: AMD Ryzen 5 5600H with Radeon Graphics (Dell G15)
• Preferred Packaging: A native .deb package would be ideal for deeper system integration, but a universally containerized format like an AppImage or Flatpak would be perfectly fine as a starting point to cover the widest range of distributions.

Agentic Workflows & Token Efficiency

Beyond convenience, running the M3 agent on Linux provides massive API token efficiency gains over Windows.

• The Windows Overhead: When the agent uses tools to execute local terminal commands, search directories, or edit files, Windows PowerShell requires verbose syntax, complex pathing, and outputs massive tables with whitespace padding that heavily inflates the context window.
• The Linux Advantage: In contrast, native Linux Bash commands and standard stdout are dense and clean. Running agentic workflows natively on Ubuntu prevents token bloat and significantly reduces API usage during multi-step tool calls.

---

Linux is the native home for most backend, cloud networking, and DevOps work. Releasing a Linux build (even just targeting Ubuntu/Wayland initially) would instantly unlock a massive demographic of power users. Happy to help test any early Alpha/Beta Linux builds!