Skip to content
Get packs

Colima

Runs Docker, containerd, or Incus containers on macOS and Linux with a simple command-line setup.

GitHub Download v0.10.3
Stars
31.5k
Forks
636
Licence
MIT
Language
Go
Last update
23 days ago
Type
Dev tool

What it is

Colima is a command-line tool that starts a lightweight Linux virtual machine on your computer and runs a container runtime inside it, such as Docker or containerd. Containers are packaged apps that run the same way across machines, and Colima lets you use them on macOS without installing a full desktop application. The README describes its goal as providing container runtimes on macOS with minimal setup.

  • Run a container with docker run after starting Colima with colima start
  • Start a containerd runtime and manage containers with nerdctl
  • Run a local Kubernetes cluster by starting Colima with --kubernetes and using kubectl
  • Launch Incus containers or virtual machines on macOS
  • Run a local AI model with colima model run on supported Apple Silicon hardware

Great for

  • Mac developers who want Docker containers without Docker Desktop
  • People who want a single command to start a container environment
  • Users who want optional Kubernetes running locally with Docker or containerd
  • Users experimenting with GPU-accelerated AI models on Apple Silicon Macs

Not ideal for

  • Windows users, since the README only lists macOS and Linux support
  • People who need Kubernetes features that require the --kubernetes flag but do not want to configure a local cluster
  • Users on Intel Macs who want the krunkit GPU features, since the README limits those to Apple Silicon and macOS 13+
Step-by-step guide: how to use ColimaInstall, first run, examples, tips and FAQs

How to get it

  1. 1

    Install with a package manager

    You'll need: Homebrew, MacPorts, Nix, or Mise (for installation) · Docker client for the Docker runtime (brew install docker) · kubectl for Kubernetes (brew install kubectl)

    Install with Homebrew

    brew install colima
  2. 2

    Run it

    • Install Colima with Homebrew using brew install colima
    • Start Colima with defaults using colima start
    • Run a test container with docker run hello-world after installing the Docker client
  3. 3

    Get the most out of it

    • Use colima start --edit to customize the VM through a config file
    • Adjust CPU, memory, and disk with flags like --cpu, --memory, and --disk
    • Use colima nerdctl install to put an nerdctl alias script in your PATH when using containerd
    • Use --vm-type=vz --vz-rosetta on Apple Silicon to run x86 images through Rosetta 2 emulation

Downloads

Source code (.zip)main branch · straight from GitHubRelease v0.10.38 files incl. colima-Linux-x86_64Official website & docscolima.run

Strengths

  • Setup is a few commands, and the README says it works with sensible defaults
  • Supports multiple runtimes (Docker, containerd, Incus) and multiple instances
  • Includes automatic port forwarding and volume mounts

Trade-offs

  • Some features have version or hardware requirements, such as Incus virtual machines needing m3 or newer Apple Silicon
  • Advanced setups rely on companion tools like kubectl, the Docker client, or krunkit that you must install yourself
  • The README is truncated here, so detailed troubleshooting lives in separate docs

How it compares

ProjectStarsTypeLevelUpdatedBest for
colima (this)31.5kDev toolbeginner23 days agoMac developers who want Docker containers without Docker Desktop
dokku32.2kAppintermediate4 days agoDevelopers who want Heroku-style deployment on a server they control
macos21.7kDev toolintermediate29 days agoDevelopers who need to test software on macOS from a Linux machine
laradock12.7kDev toolintermediate6 days agoPHP developers who want a consistent local environment across Linux, macOS and Windows
dockge24.7kAppbeginner14 days agoPeople who run Docker Compose stacks on a home server or small Linux machine
lens23.2kAppbeginner1.7 years agoDevelopers who want a visual view of their Kubernetes workloads

Stats synced from GitHub. The assessment above was drafted with AI from the project's README on 2026-10-29 and may miss recent changes — check the README before relying on a command. MIT licence: free to use, modify and distribute, including commercially, as long as you keep the copyright notice.