← Downloads

Incus 7.3

Package v7.3 x86_64 TAR.GZ July 31, 2026

A system container and virtual machine manager forked from LXD under the Linux Containers project, managing both with one CLI, storage pools, networking, snapshots, and clustering.

Download TAR.GZ Project website ↗

Download Mirrors

Mirror Region Download
Incus (Official) Primary Global Download
Incus Downloads Global Download
GitHub Repository Global Download

Incus manages system containers and virtual machines with the same commands. It is a community fork of LXD, created after Canonical moved LXD under a contributor licence agreement, and it is maintained under the Linux Containers project alongside LXC.

System containers, not application containers

The distinction matters and is frequently missed.

A Docker container runs one process and finishes when that process exits. A system container boots an init system and runs a full userspace with services, users, cron, and logins. It behaves like a small machine and starts in about a second.

Both use the same kernel primitives. The difference is entirely what runs inside. Our LXC and Incus guide covers the model.

Using it

sudo apt install incus        # Debian and Ubuntu
sudo dnf install incus        # Fedora
sudo pacman -S incus          # Arch

sudo incus admin init                    # interactive first-time setup
sudo usermod -aG incus-admin "$USER"     # then log out and back in
incus launch images:debian/13 web
incus exec web -- bash
incus config set web limits.memory 2GiB
incus snapshot create web before-upgrade
incus stop web && incus delete web

incus exec web -- bash is the moment the model clicks: a root shell in what feels like a fresh Debian install, with systemd running, that started instantly.

It runs virtual machines too

incus launch images:debian/13 vm1 --vm

Same commands, same storage, same networking, but a real VM with its own kernel. That flexibility is the practical argument for Incus over running containers and libvirt separately.

Storage backend matters

incus admin init asks about storage, and the answer has real consequences. On ZFS or Btrfs, snapshots and clones are near-instant because they are copy-on-write. On the dir backend they work and are slow copies.

Choose deliberately at init; changing later means recreating the pool.

Unprivileged by default

Containers run unprivileged, mapping root inside to an unprivileged host user. A container escape lands in a nobody account rather than as host root.

That is a large improvement over privileged containers and it is not a virtual machine boundary, because the kernel is shared. For genuinely untrusted workloads, use the VM mode.

Verify Your Download

Most installations come from distribution packages or the Zabbly repository the project maintains. Source releases are on GitHub.