Install Kortix, the open-source AI Operating System, in three commands

What you need

You need a Mac or a Linux machine, and a terminal. The Kortix CLI runs on macOS and Linux, and on Windows through WSL: there is no Windows CLI binary and no Chrome extension to install.

The managed-cloud path needs nothing beyond that. The self-host path needs a Linux box with Docker, plus a domain for production or a Cloudflare tunnel for evaluation. On a bare Linux box, the one-shot script in the self-hosting docs installs Docker, installs the Kortix CLI, and starts the stack in one command. It runs on Linux only.

Three commands to your first session

1. curl -fsSL https://kortix.com/install | bash installs the Kortix CLI.
2. kortix init scaffolds a project. It creates kortix.yaml plus your agents, skills, and runtime config.
3. kortix ship checks the project, asks for missing secrets, pushes your repo, and brings the whole thing live in the cloud.

Start work, then review what comes back:

kortix sessions new --prompt "Summarize this week's commits and open a change request"kortix cr ls   # review what an agent proposes, merge to keep itkortix chat    # talk to a session's agent from your terminal

Every change request gets a preview you can open, and merge is default-deny for agents: you review the diff before anything lands on main. For a walkthrough of the first session, read the getting-started guide.

What the first run asks

On a self-hosted instance, the first command after the stack starts is kortix self-host configure. It opens an interactive prompt for your sandbox provider key, and optionally a managed-git token. Agent sessions run on that separate sandbox provider. The default is Daytona, and Platinum and E2B are also supported.

Models come from you, in the app. Sign up in the dashboard, then connect your own LLM key in the model picker. Self-hosted instances use your own key by default: any model provider with your own keys, or the ChatGPT plan you already pay for.

Session work reaches main through a change request a person approves. Approval gates are off until you set them.

Where to run it

Kortix runs on Kortix Cloud, in your VPC, or on your own on-prem network. Self-host is free.

Kortix Cloud. Nothing to install. Sign up at kortix.com, create a project, and start a session.

Your VPC. The same stack, inside your own cloud account.

On-prem. Your own box runs one Docker Compose stack: the frontend, the API, the LLM gateway, and the Supabase distribution. After the CLI install, the manual path is:

1. Point DNS at the box: an A/AAAA record for your domain and for api.<domain>, both pointing at the box IP. Open ports 80 and 443 so the bundled Caddy proxy can issue a TLS certificate.
2. Run kortix self-host init --domain kortix.example.com.
3. Run kortix self-host start, then check status, logs, and doctor while the stack starts.

For an evaluation with no domain, run kortix self-host init --tunnel cloudflare instead. The tunnel URL changes on every restart, so use that mode for evaluation only.

Updates install automatically. Pin an exact version with kortix self-host update --tag 0.9.84, or turn the updater off with --auto-update off. Before any destructive command, back up the database directory, the file storage directory, and the instance .env file under ~/.config/kortix/self-host/<instance>/.

Bringing an existing setup with you? Read the migration guide. The full host setup lives in the self-hosting guide on this site, and you can read the self-hosting docs for every kortix self-host subcommand.