How to Install OpenClaw on Your Windows PC or Macbook [2026]

Why Set Up OpenClaw on Windows? If you have been looking for a way to run a powerful, self-hosted AI assistant right on your own Windows machine, you are in the right place. OpenClaw is an open-source AI assistant framework that gives you full control, privacy, and flexibility. Instead of relying on someone else’s servers, OpenClaw runs locally on your laptop or desktop, connecting to the AI models you choose and the messaging platforms you already use.

So what can you actually do with OpenClaw once it is set up? A lot. You can chat with it through Telegram, Discord, Slack, Signal, or even a web interface. You can teach it skills to automate tasks, search the web, manage your calendar, or control smart home devices. You can customise how it thinks, speaks, and responds. And because it runs on your own hardware, your conversations stay on your machine unless you explicitly choose to share them.

This guide walks you through the complete setup process on Windows. We cover three different installation paths, help you pick the right one for your situation, and make sure you are not left scratching your head when something does not work as expected. By the end of this guide, you will have a fully running OpenClaw instance that greets you every time you open your laptop.

System Requirements: What You Need Before You Start

Before you run a single command, let us make sure your machine can handle OpenClaw. The good news is that the requirements are pretty reasonable for a modern Windows PC or laptop.

Here is what you need:

Internet Connection
You need an internet connection to download OpenClaw, connect to messaging platforms, and communicate with AI model providers.

That is pretty much it. If your laptop is from the last five years and runs Windows 10 or 11 without crawling, you are good to go.

Choosing Your Installation Path: WSL2 vs. Native Windows vs. Docker

This is the most important decision you will make before installing. OpenClaw supports three different paths on Windows, and each one has its own personality. Here is a straightforward breakdown to help you decide.

WSL2 (Windows Subsystem for Linux 2) — Recommended

WSL2 lets you run a full Linux environment inside Windows without dual-booting. Think of it as a lightweight virtual machine that is deeply integrated into Windows. OpenClaw was originally built for Linux and macOS, so running it inside WSL2 gives you the smoothest, most stable experience. You get full access to all OpenClaw features, the CLI, the Gateway, and all available tooling without any compatibility headaches.

Choose WSL2 if you want the full OpenClaw experience, you are comfortable spending 15 extra minutes setting it up, or you plan to use OpenClaw seriously as your daily assistant.

Learn more about WSL2 and how to enable it directly from Microsoft at https://learn.microsoft.com/en-us/windows/wsl/install.

Native Windows (PowerShell) — Good for CLI-Only Users

You can also install OpenClaw directly on Windows using PowerShell without any Linux layer. This works well for core CLI usage. You can run commands like openclaw --version, openclaw doctor, and openclaw plugins list. The gateway and basic assistant functions also work. However, some advanced features and tooling have minor limitations compared to the WSL2 path.

Choose native Windows if you just want to experiment with OpenClaw, you do not need every single feature, or you are setting up a quick test environment.

Docker Desktop — Best for Container Fans

If you already use Docker and love working with containers, OpenClaw has an official Docker image you can pull and run. This keeps OpenClaw isolated from your host system and makes updates and rollbacks very clean.

Choose Docker if you are already using Docker Desktop, you want easy cleanup, or you need to run multiple OpenClaw instances side by side.

You can find the full Docker setup guide at https://docs.openclaw.ai/install/docker.

Method 1: Installing OpenClaw via WSL2 (Recommended)

This is the gold standard for Windows users. It takes a bit more time but pays off in stability and feature coverage. Follow these steps in order and you will be up and running.

Step 1: Enable WSL2 on Your Windows Machine

First, you need to turn on the Windows Subsystem for Linux feature and install WSL2. Open PowerShell as Administrator (right-click the Start menu, choose “Terminal (Admin)” or “PowerShell (Admin)”) and run:

wsl --install

This command enables all the required Windows features and installs Ubuntu as your Linux distribution. After it finishes, restart your computer.

If you already have WSL installed but are not sure whether it is WSL1 or WSL2, run this command in PowerShell:

wsl --set-default-version 2

This ensures new distributions use WSL2. For more details, Microsoft’s official WSL2 documentation at https://learn.microsoft.com/en-us/windows/wsl/ is an excellent reference.

Step 2: Set Up Your Linux Distribution

After restarting, WSL will launch automatically and ask you to create a Linux username and password. Pick anything you like, just make sure you remember the password since you will need it for admin tasks.

Once you are inside the Linux terminal, run a quick update:

sudo apt update && sudo apt upgrade -y

This makes sure your Linux environment is fully up to date before installing anything new.

Step 3: Install OpenClaw Using the Official Script

Here is the moment you have been waiting for. The OpenClaw team provides a one-command installer that handles everything automatically. In your WSL2 terminal, run:

curl -fsSL https://openclaw.ai/install.sh | bash

The script is clever. It detects your operating system, checks whether Node.js is installed, and installs it if needed. Then it pulls down OpenClaw and sets it up. This typically takes two to five minutes depending on your internet speed.

If you prefer not to go through the onboarding prompts automatically, you can run the installer with a flag to skip that step:

curl -fsSL https://openclaw.ai/install.sh | bash -s -- --no-onboard

You can always run the onboarding manually later.

Step 4: Connect Your AI and Set Up the Gateway

After the installer finishes, you need to connect OpenClaw to an AI model provider. OpenClaw supports a range of options. The easiest way to get started is to run the onboarding wizard:

openclaw onboard --install-daemon

This walks you through setting up your API key, choosing a model provider, and configuring the gateway. The gateway is the background service that keeps OpenClaw running and accessible.

If you are not sure which model provider to choose, OpenClaw works well with OpenAI-compatible APIs, local models, and several other options. The official documentation at https://docs.openclaw.ai/start/getting-started has a detailed breakdown of each option with pros and cons.

Step 5: Configure Auto-Start with Systemd

You want OpenClaw to start automatically every time you boot your laptop, right? Inside WSL2, you can set that up using systemd. Run:

sudo loginctl enable-linger "$(whoami)"

This tells the system to keep your user services running even when you are not logged in. Then install the gateway as a user service:

openclaw gateway install

Now your OpenClaw assistant will always be running in the background, ready the moment your laptop wakes up.

Step 6: Verify Everything Is Working

Run these three commands to confirm everything is set up correctly:

openclaw --version
openclaw doctor
openclaw gateway status

The first confirms the CLI is installed. The second checks your configuration for any issues. The third shows whether the gateway is running. If all three give you green lights, you are done.

Extra Tip: Want Your Assistant Running Before Windows Login?

If you are running a headless setup or a server, you can configure OpenClaw to start even before the Windows login screen appears. The official docs at https://docs.openclaw.ai/platforms/windows cover this advanced setup in detail, including using Windows Scheduled Tasks to boot WSL2 automatically at startup.

Method 2: Installing OpenClaw Directly on Windows (PowerShell)

If WSL2 feels like too much setup, or you just want to get OpenClaw running quickly for testing, the native Windows path is a perfectly valid choice. Here is how to do it.

Step 1: Install Node.js

OpenClaw needs Node.js. The easiest way to manage Node versions on Windows is to use nvm (Node Version Manager) for Windows. Download it from https://github.com/coreybutler/nvm-windows/releases and run the installer.

Once installed, open a new PowerShell window (close any old ones) and install the latest Node.js:

nvm install 24
nvm use 24

Alternatively, you can download Node.js directly from https://nodejs.org/ and install it the traditional way.

Step 2: Run the Windows Installer Script

Open PowerShell as Administrator. Yes, always as Administrator for this step. Then run:

iwr -useb https://openclaw.ai/install.ps1 | iex

The script downloads OpenClaw and sets it up on your Windows system. If you encounter any permission errors, make sure your PowerShell window is running with administrator privileges.

To skip the interactive onboarding prompts:

& ([scriptblock]::Create((iwr -useb https://openclaw.ai/install.ps1))) -NoOnboard

Step 3: Connect Your API Key and Start the Gateway

Run the onboarding wizard:

openclaw onboard --install-daemon

On Windows, this tries to set up a Scheduled Task for auto-start. If your system denies Scheduled Task creation (some work computers restrict this), OpenClaw automatically falls back to a Startup folder shortcut, which still runs the gateway every time you log in.

Step 4: Verify the Installation

openclaw --version
openclaw doctor
openclaw gateway status

You should see the version number, a clean bill of health from the doctor command, and the gateway status showing it is running.

A Few Things to Know About the Native Windows Path

The OpenClaw team notes that some advanced CLI tooling and certain system integrations are still being improved for native Windows. The WSL2 path remains more comprehensive. That said, if you just want the core assistant running with a web interface or a messaging channel like Telegram, the native path works great.

For the latest status on native Windows compatibility, check the official Windows platform page at https://docs.openclaw.ai/platforms/windows.

Method 3: Installing OpenClaw with Docker Desktop

If you are already in the Docker ecosystem, this is the cleanest and most portable way to run OpenClaw on Windows.

First, install Docker Desktop from https://www.docker.com/products/docker-desktop/ if you do not already have it.

Then pull the official OpenClaw image and run it:

docker pull openclaw/openclaw
docker run -d --name openclaw -p 18789:18789 openclaw/openclaw

This starts OpenClaw in detached mode and maps the gateway port. You can then access it through your browser or connect your messaging channels.

For a more complete setup with persistent configuration and volumes, the official Docker documentation at https://docs.openclaw.ai/install/docker walks you through environment variables, port configuration, and Docker Compose examples.

Post-Install Setup

Now that OpenClaw is installed, let us make it useful. Here is what most users do next.

Running the Doctor Check

Always start with openclaw doctor. This command inspects your configuration, checks that all dependencies are in place, and flags anything that needs your attention. Think of it as a health checkup for your assistant.

Connecting Messaging Channels

One of OpenClaw’s superpowers is connecting to platforms you already use. You can link Telegram, Discord, Slack, Signal, WhatsApp, and more. The official setup for each channel varies slightly. You can find platform-specific guides at https://docs.openclaw.ai/channels.

The general process involves creating a bot token (for Telegram, Discord, etc.), configuring the channel in your OpenClaw config, and restarting the gateway. It sounds more complicated than it is, and the docs make it straightforward.

Installing Skills from ClawHub

OpenClaw has a growing library of pre-built skills that extend what your assistant can do. Things like web search, calendar management, code execution, image generation, and much more. You can browse and install skills from ClawHub at https://clawhub.ai.

To install a skill, you use the ClawHub CLI:

clawhub install <skill-name>

The skill documentation tells you exactly what command to run.

Configuring Your AI Model

By default, OpenClaw will ask you to provide an API key during onboarding. OpenClaw is model-agnostic, meaning it works with any LLM provider that speaks the OpenAI-compatible API format. This includes OpenAI itself, Anthropic (with a proxy layer), local models via LM Studio or Ollama, and many others.

If you want to run entirely offline with local models, check out https://docs.openclaw.ai/providers/local for a step-by-step guide using LM Studio or Ollama. Running local models is slower than using a cloud API, but it keeps everything 100 per cent on your machine.

Troubleshooting Common Issues

Even with the best installer, things can sometimes go sideways. Here is a quick reference for the most common problems Windows users encounter.

Problem: “openclaw is not recognised as an internal or external command”

This is the most common issue by far. It means the system cannot find the OpenClaw executable, usually because the PATH environment variable was not updated correctly after installation.

To fix it, find where OpenClaw is installed. In PowerShell or Command Prompt, run:

npm prefix -g

This shows you the global npm directory. Make sure the bin the folder inside that path is in your system PATH. On Windows, go to Settings > System > About > Advanced system settings > Environment Variables, find “Path” under System variables, and add the bin folder path if it is missing.

After fixing the PATH, close and reopen your terminal window for the changes to take effect.

Problem: Gateway will not start

First, run openclaw doctor to see what it reports. Common causes include a port conflict (something else using port 18789), missing Node.js, or a misconfigured API key.

If the doctor shows no obvious issues, try running the gateway manually to see any error messages:

openclaw gateway run

This will output any errors directly to the terminal so you can see what is going wrong.

Problem: Scheduled Task creation is denied

Some Windows machines, especially those managed by schools or employers, restrict Scheduled Task creation.

If openclaw onboard --install-daemon fails at this step, it is not the end of the world. OpenClaw automatically falls back to creating a shortcut in your user Startup folder. The gateway will still start every time you log in.

For a workaround using Windows Task Scheduler manually, the troubleshooting section at https://docs.openclaw.ai/platforms/windows has specific commands you can try.

Problem: WSL2 feels slow or does not start

Make sure Virtual Machine Platform is enabled in Windows Features. Open Control Panel > Programs > Turn Windows Features on or off and check “Virtual Machine Platform” and “Windows Subsystem for Linux.” Restart your machine after enabling these.

If WSL2 is still slow, make sure you are on WSL2 and not WSL1:

wsl --list --verbose

If a distribution shows version 1 next to it, upgrade it:

wsl --set-version <distro-name> 2

Keeping OpenClaw Updated

OpenClaw is actively developed, with updates released regularly. Here is how to stay current.

To update OpenClaw when installed via the installer script, you can reinstall:

curl -fsSL https://openclaw.ai/install.sh | bash

If you installed via npm:

npm install -g openclaw@latest

The @latest tag always pulls the most recent stable release.

To check your current version without updating:

openclaw --version

And if you ever need to start fresh or remove OpenClaw completely, the uninstall guide at https://docs.openclaw.ai/install/uninstall walks you through cleaning up all files and services.

Conclusion

Setting up OpenClaw on a Windows laptop or desktop is a genuinely rewarding project. Once it is running, you have a private, powerful AI assistant that works on your schedule, connects to your tools, and learns how you like to work.

Here is a quick recap of the three paths:

Start with WSL2 unless you have a specific reason not to. The extra 15 minutes of setup time is worth it.

After you have OpenClaw running, the next best step is to connect at least one messaging channel so you can actually talk to your assistant from your phone or another device. Then browse ClawHub and install a skill or two that match what you want to do. You will be surprised how quickly it becomes part of your daily workflow.

Good luck, and enjoy your new AI assistant.