How to install OpenClaw
Install the OpenClaw personal agent on macOS, Linux, or Windows: installer script, npm, source builds, and the onboarding wizard explained step by step.
OpenClaw is an open-source personal AI agent that connects to your messaging apps and takes real actions on your behalf: running commands, browsing the web, managing files, and more. It runs on your own machine, which means you install and operate it yourself. This guide covers every supported install path and what to expect from onboarding.
Prerequisites
- Node 24 (recommended) or Node 22.19+. The installer script handles Node for you, so you only need this if you install via a package manager.
- macOS, Linux, or Windows. On Windows you can use the native Windows Hub app, the PowerShell installer, or WSL2.
pnpmonly if you build from source.
Recommended: the installer script
The fastest path detects your OS, installs Node if needed, installs OpenClaw, and launches onboarding.
macOS / Linux / WSL2:
curl -fsSL https://openclaw.ai/install.sh | bash
Windows (PowerShell):
iwr -useb https://openclaw.ai/install.ps1 | iex
To install without launching onboarding, append --no-onboard:
curl -fsSL https://openclaw.ai/install.sh | bash -s -- --no-onboard
Install with a package manager
npm:
npm install -g openclaw@latest
openclaw onboard --install-daemon
pnpm:
pnpm add -g openclaw@latest
pnpm approve-builds -g
openclaw onboard --install-daemon
bun:
bun add -g openclaw@latest
openclaw onboard --install-daemon
Build from source
git clone https://github.com/openclaw/openclaw.git
cd openclaw
pnpm install && pnpm build && pnpm ui:build
pnpm link --global
openclaw onboard --install-daemon
What onboarding does
The openclaw onboard wizard takes about two minutes and walks you through:
- Choosing a model provider (Anthropic, OpenAI, Google, and others). Not sure which model to run? See our best models for OpenClaw ranking.
- Entering an API key.
- Configuring the Gateway, the always-on backend service that handles model connections and message routing. The Gateway runs on port 18789 and must be active before chat works.
The --install-daemon flag registers the Gateway as a managed service: LaunchAgent on macOS, systemd on Linux/WSL2, or a Scheduled Task on Windows, so it survives reboots.
Verify the install
openclaw --version # confirm the CLI is available
openclaw doctor # check for config issues
openclaw gateway status # verify the Gateway is running
openclaw dashboard # open the Control UI in your browser
Send your first message from the Control UI dashboard. After that, connect a messaging channel (Telegram is the fastest, needing only a bot token) and chat with your agent from anywhere.
Before you expose it to the world
OpenClaw can run shell commands and read your files, so harden it before connecting public channels. Read our guide to running OpenClaw safely for the security audit command, pairing policies, and sandboxing.
Source: OpenClaw install docs and getting started guide, checked June 2026.