How to Deploy OpenClaw on a Windows PC — A Simple Tutorial

OpenClaw has been getting a lot of attention lately. If 2025 was the year DeepSeek’s models went mainstream, 2026 is clearly the year OpenClaw is making AI agents popular.

You may have already seen people “running OpenClaw” in various places — that usually means they’ve installed and deployed OpenClaw on their computer.

What is OpenClaw?

OpenClaw is an AI tool you can run on your own PC. Think of it as a smart controller installed on your machine.

Once it’s there, you can talk to it and use that conversation to control apps, programs, and workflows on your computer — in other words, let the AI do work for you.

OpenClaw was created by Peter Steinberger, a software engineer from Austria. It’s an open-source AI agent framework, so OpenClaw itself is free and open source.

From experience, installation isn’t especially hard. If you’re comfortable with basic computer use and have installed something like Python before, deploying OpenClaw is manageable.

Before we get to the actual steps, here are a few important points.

Before You Start: Things You Should Know

- OpenClaw is free, but using it requires an AI model API. You need an API account (e.g. OpenAI, Anthropic, DeepSeek, etc.) and keys.

- Running agents with OpenClaw uses tokens. Token usage is often paid. Be aware of your provider’s pricing.

- OpenClaw can start apps and read files on your computer when it’s doing automated tasks. Only run it if you’re comfortable with that.

- Installation needs administrator rights on your PC. That can introduce security and account risks if you’re not careful.

So: before you deploy OpenClaw on your main machine, understand these trade-offs.

If you can, use a spare or test PC so that if something goes wrong, the impact is limited.

Do You Need a High-End PC?

No. For deployment itself you don’t need powerful hardware:

- RAM: 4 GB or more  

- Disk: At least 500 MB free  

- OS: Windows 10 is enough (e.g. a typical office PC)

Overview: Three Main Steps

Deploying OpenClaw on your PC boils down to:

1. Check and install the dependency environment (Node.js, npm, Git)  

2. Install OpenClaw

3. Set up the AI conversation channel (model API + channels like WebChat, etc.)

You can either install everything locally on your PC (more control and privacy) or use cloud-based one-click options that some vendors offer (easier, but data goes to the cloud). This tutorial focuses on local installation on Windows.

Step 1: Check and Install the Dependency Environment

OpenClaw needs Node.js, npm, and Git on your machine.

Check if they’re already installed

1. Press Win + R to open the Run dialog.  
2. Type cmd and press Enter to open Command Prompt.  

3. Run:

node -v

then:

npm -v

If you see version numbers (e.g. Node v22.x, npm 10.x), you’re good. OpenClaw expects Node 22 or higher. If either command fails or the version is too old, install or upgrade as below.

Install Node.js (if needed)

1. Go to: https://nodejs.org/ (or the [downloads page](https://nodejs.org/en/download/) for your OS).  

2. Download the Node 22+ installer for Windows.  

1. Run the installer and follow the steps (you can accept defaults, or customize if you know what you need).
2. Close and reopen Command Prompt, then run node -v and npm -v again to confirm.

Install Git (if needed)

1. Download the Windows Git installer from: https://git-scm.com/download/win.  

2. Run the installer.

3. Go through the wizard (default options are fine). Click Next until it finishes, then Finish.  

4. Open a new Command Prompt and run:

git --version

to confirm Git is installed.

Once Node.js, npm, and Git all show version numbers, your dependency environment is ready. Move on to Step 2.

Step 2: Install OpenClaw

1. Click the Start menu (bottom-left), find Windows PowerShell, and open it.  

2. Copy the following command, paste it into the blue PowerShell window, and press Enter:

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

3. Wait for the script to run. When you see a success message (e.g. “Installation complete” or similar), OpenClaw is installed.

At this point OpenClaw is on your PC, but you still can’t use it for AI chat until you add an LLM and a conversation channel. That’s Step 3.

Step 3: Set Up the AI Conversation Channel

You need to:

- Configure an LLM (large language model) with your API key.  

- Configure at least one channel (the way you talk to OpenClaw, e.g. WebChat in the browser, or integrations like Feishu, WeChat, etc.).

Run the onboarding wizard

1. Open Command Prompt or PowerShell.
2. Run:

openclaw onboard --install-daemon

3. When prompted, choose Yes to continue.  

4. The wizard will show model providers (e.g. OpenAI, Anthropic, DeepSeek, etc.). Pick the one you have an API key for.  

5. Enter your API key when asked.  

6. Select the model version you want (e.g. a specific Claude or GPT model).  

7. Then configure a channel. The channel is how you send messages to OpenClaw — for example:

  - WebChat (built-in web UI), or  

  - Integrations like Feishu, WeChat, QQ, etc., if you install and configure them.  

8. Follow the on-screen prompts until the wizard finishes.

After this, OpenClaw has an LLM and a channel. You can start giving it tasks via the chosen app or the web interface.

Start OpenClaw and Open the Web UI

1. In a terminal, run:

openclaw gateway

2. Leave that window open (the gateway is now running).

3. In your browser, go to: http://127.0.0.1:18789 (this is the default address).  

4. You should see the OpenClaw web interface (dashboard). If you’ve set an API key and model, the gateway will show as online; otherwise it may show offline until the model is configured.

You’re Done

After completing:

- Step 1: Dependencies (Node.js, npm, Git)  

- Step 2: OpenClaw installation  

- Step 3: LLM API key + channel (and optionally `openclaw onboard --install-daemon` for a background service)

you have OpenClaw deployed on your Windows PC. You can start the gateway with `openclaw gateway`, open http://127.0.0.1:18789 in your browser, and use the web UI or your configured channels to talk to OpenClaw and let it help you run apps and automate tasks on your machine.

For more options (e.g. different install methods, plugins, or security), see the official OpenClaw documentation.