How to Install OpenClaw for Complete Beginners: Step-by-Step Guide

Quick Answer: To install OpenClaw, you need Node.js 22+, then run the installation script for your operating system (macOS/Linux: curl -fsSL https://openclaw.ai/install.sh | bash or Windows via WSL2: iwr -useb https://openclaw.ai/install.ps1 | iex), followed by the onboarding wizard (openclaw onboard) to configure your AI model, API key, and messaging channels.

What is OpenClaw and Why Should You Install It?

OpenClaw is an open-source personal AI assistant that runs entirely on your own computer. Unlike cloud-based assistants that send your data to remote servers, OpenClaw keeps everything local while giving you AI-powered help through the messaging apps you already use every day.

Think of OpenClaw as your own private AI helper that can manage your calendar, browse the web for information, organize files, and run commands on your computer—all while respecting your privacy. You interact with it through WhatsApp, Telegram, Discord, Slack, or even iMessage, making it feel natural and convenient.

The project has exploded in popularity with over 160,000 GitHub stars, becoming the go-to self-hosted AI assistant for people who value privacy and control over their data. When you install OpenClaw, you're joining a massive open-source community that's building the future of personal AI without sacrificing privacy.

Here's what makes OpenClaw worth installing:

• Complete privacy: Your data never leaves your computer unless you explicitly tell it to

• Messaging app integration: Chat with your AI assistant using apps you already know

• AI model flexibility: Works with Claude, GPT, DeepSeek, or any major AI model

• Powerful automation: Handles tasks like file management, web research, and system commands

• Open source transparency: You can see exactly what the code does—no hidden surprises

The installation process takes about 20 minutes for complete beginners, and you'll have your own AI assistant up and running by the end of this guide.

What Are the System Requirements for Installing OpenClaw?

Before you start the installation, make sure your computer meets these requirements. OpenClaw is surprisingly lightweight, but it needs specific versions of certain software to work correctly.

Operating System Requirements

macOS Users:

• macOS 13 (Ventura) or newer

• Works on both Intel and Apple Silicon (M1/M2/M3/M4) chips

• About 500 MB of disk space for installation

• At least 5 GB of free space recommended for long-term use

Linux Users:

• Any modern Linux distribution works (Ubuntu, Debian, Fedora, Arch, etc.)

• 500 MB for the initial installation

• 5 GB of free disk space for comfortable operation

Windows Users:

• Windows 10 or Windows 11

• Important: OpenClaw doesn't run natively on Windows

• You must use WSL2 (Windows Subsystem for Linux) with Ubuntu

• This gives you a Linux environment inside Windows where OpenClaw can run properly

Why doesn't OpenClaw support native Windows? The software relies on Unix-based process management and protocols that expect a POSIX environment. Features like WhatsApp Web integration and iMessage support are built for Unix systems, making WSL2 the only reliable path for Windows users.

Software Prerequisites

Node.js Version:

• Node.js 22 or newer is required

• npm comes automatically with Node.js (you don't need to install it separately)

• Older Node versions will cause installation failures or crashes under load

Why does OpenClaw need such a recent Node.js version? The software uses modern JavaScript features and up-to-date security patches that only exist in newer Node releases. Using an older version creates instability.

Hardware Requirements:

• 8 GB of RAM for basic use with cloud AI models

• 16 GB or more recommended for heavy automation workloads

• Any modern processor from the last 5 years will work fine

AI Model Access:

• You'll need an API key from at least one AI provider

• Anthropic (Claude) is recommended by the OpenClaw team

• Alternative options: OpenAI (GPT), DeepSeek, or other compatible models

• Budget roughly $5-20/month for typical usage (depends on how much you use the assistant)

Comparison: OpenClaw vs Common Alternatives

How Do You Install OpenClaw on macOS and Linux?

The installation process for macOS and Linux is straightforward and mostly automated. The installer script handles everything from downloading files to setting up the initial configuration.

Step 1: Install Node.js 22 or Newer

Before running the OpenClaw installer, you need Node.js installed on your system.

On macOS: The easiest way is using Homebrew, a popular package manager for Mac.

# Install Homebrew first if you don't have it

/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)"

# Install Node.js

brew install node@22

On Linux: Most Linux distributions have Node.js in their package repositories, but they might not have version 22 yet. Here's how to get the latest version.

# Ubuntu/Debian

curl -fsSL https://deb.nodesource.com/setup_22.x | sudo -E bash -

sudo apt-get install -y nodejs

# Fedora

sudo dnf install nodejs

# Arch Linux

sudo pacman -S nodejs npm

Verify your Node.js version by running:

node --version

You should see something like v22.x.x. If the version is older than 22, you'll need to update before continuing.

Step 2: Run the OpenClaw Installation Script

Now comes the easy part. Run this command in your terminal:

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

This command downloads and runs the official OpenClaw installation script. The script automatically:

• Detects your operating system

• Downloads the latest OpenClaw version

• Installs it globally on your system

• Sets up the necessary file permissions

• Configures your system PATH so you can run openclaw commands

The installation takes 2-5 minutes depending on your internet speed. You'll see progress messages as it works.

Step 3: Run the Onboarding Wizard

After installation completes, launch the interactive setup wizard:

openclaw onboard

The onboarding wizard is where all the important configuration happens. It walks you through several setup steps with clear prompts and helpful explanations.

Here's what the wizard will ask you:

1. Risk Acknowledgment: OpenClaw is powerful experimental software that can access your files and run commands. You'll see a warning explaining the risks. Read it carefully, then use the arrow keys to select "Yes, I understand" if you want to proceed.

2. Installation Type: Choose "QuickStart" for the recommended default settings. This option is perfect for beginners because it makes sensible choices for you.

3. AI Model Selection: Select which AI model you want to use. Anthropic (Claude) is recommended because the OpenClaw team optimizes for it, but you can choose OpenAI, DeepSeek, or others.

4. API Key Configuration: You'll need to provide your API key. The wizard gives you clear instructions on where to get one (we'll cover this in detail in the next section).

5. Messaging Channel Setup: Choose which messaging app you want to connect. Start with just one—you can add more later. Each channel has its own setup process that the wizard guides you through.

The onboarding wizard saves all your settings to ~/.config/openclaw/openclaw.json5, which is the main configuration file.

Step 4: Start the OpenClaw Gateway

The gateway is the core service that keeps OpenClaw running and handles communication between your AI model and messaging apps. To verify OpenClaw is working, you can start it manually or let it run in the background as a system service.

The daemon mode (service) is what most people want because OpenClaw keeps running even after you close your terminal.

After onboarding completes, the wizard typically starts the gateway automatically. You can check if it's running with:

openclaw status

If you need to start it manually:

openclaw gateway start

You'll see a message confirming the gateway is running, and you can access the web dashboard at http://localhost:18789.

How Do You Install OpenClaw on Windows Using WSL2?

Windows users need an extra step because OpenClaw requires a Unix-like environment. WSL2 (Windows Subsystem for Linux) gives you a real Linux system running inside Windows, which is exactly what OpenClaw needs.

Step 1: Install WSL2 with Ubuntu

Open PowerShell as Administrator (right-click the Start menu and choose "Windows PowerShell (Admin)"), then run:

wsl --install

This single command installs WSL2 and Ubuntu automatically. It's the official Microsoft method and handles all the complex setup for you.

You'll need to restart your computer after installation. When Windows boots back up, Ubuntu will finish its setup automatically and ask you to create a username and password for your Linux environment.

Important: Remember this password—you'll need it whenever you run sudo commands inside Ubuntu.

Step 2: Open Ubuntu and Update the System

After restart, search for "Ubuntu" in your Start menu and open it. You'll see a Linux terminal.

Update your package lists to make sure you're getting the latest software versions:

sudo apt update

sudo apt upgrade -y

This ensures your Ubuntu installation has all the latest security patches and software versions.

Step 3: Install Node.js 22 in Ubuntu

Inside your Ubuntu terminal, install Node.js:

curl -fsSL https://deb.nodesource.com/setup_22.x | sudo -E bash -

sudo apt-get install -y nodejs

Verify the installation:

node --version

Step 4: Install OpenClaw

Now you can follow the same installation process as Linux users:

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

Then run the onboarding wizard:

openclaw onboard

Everything from this point forward works exactly like it does on macOS and Linux. Your OpenClaw instance runs inside the Ubuntu environment, but you can access it from Windows through the web dashboard or messaging apps.

How Do You Configure Your AI Model and API Key?

OpenClaw doesn't include AI capabilities on its own—it connects to external AI models through their APIs. You need to choose a provider and set up an API key so OpenClaw can access the AI.

Choosing Your AI Provider

OpenClaw supports multiple AI providers. Here are the most popular options:

Anthropic (Claude) - Recommended

• Models: Claude 3.5 Sonnet, Claude 3 Opus, Claude 3 Haiku

• Strengths: Excellent at following complex instructions, strong reasoning, good value

• Cost: Pay-per-use, roughly $3-15/month for typical personal use

• API dashboard: console.anthropic.com

OpenAI (GPT)

• Models: GPT-4, GPT-4 Turbo, GPT-3.5

• Strengths: Very capable, widely supported

• Cost: Similar to Anthropic for comparable models

• API dashboard: platform.openai.com

DeepSeek

• Models: DeepSeek V2, DeepSeek Coder

• Strengths: Cost-effective, good for coding tasks

• Cost: Generally cheaper than Anthropic/OpenAI

• API dashboard: platform.deepseek.com

For beginners, Anthropic's Claude is the best choice because OpenClaw's development team tests against it most frequently, which means fewer edge-case bugs.

Getting Your API Key (Anthropic Example)

Here's the detailed process for getting an Anthropic API key:

1. Create an Anthropic Account:

• Go to console.anthropic.com in your web browser

• Click "Sign Up" and create an account with your email

• Verify your email address

2. Add Billing Information:

• Navigate to Settings → Billing

• Add a credit or debit card

• Anthropic requires payment info before you can use the API

• Don't worry—you only pay for what you use, and they notify you before charges

3. Generate Your API Key:

• Go to Settings → API Keys

• Click "Create Key"

• Give it a name like "OpenClaw Main"

• Copy the key immediately—it starts with sk-ant-api03-

• Important: You can't see the key again after you close the window, so save it somewhere secure

4. Set a Spending Limit:

• In the Billing section, set a monthly spending limit

• Start with $20/month to avoid surprise bills while you're learning

• You can adjust this later based on your actual usage

Configuring the API Key in OpenClaw

If you ran the onboarding wizard, you already entered your API key during that process. But if you need to change it later or add a different provider, here's how:

Method 1: Re-run the Onboarding Wizard

openclaw onboard

Select the provider you want to configure and enter the new API key when prompted.

Method 2: Manual Configuration

Open the configuration file in a text editor:

nano ~/.config/openclaw/openclaw.json5

Find the models.providers section and add or update your key:

{

  models: {

    providers: {

      anthropic: {

        apiKey: "sk-ant-api03-YOUR-KEY-HERE"

      }

    }

  }

}

Save the file and restart the gateway:

openclaw gateway restart

Method 3: Environment Variable

You can also set the key as an environment variable:

export ANTHROPIC_API_KEY="sk-ant-api03-YOUR-KEY-HERE"

Add this line to your ~/.bashrc or ~/.zshrc file to make it permanent.

How Do You Connect OpenClaw to WhatsApp, Telegram, or Discord?

One of OpenClaw's best features is that you can chat with your AI assistant using messaging apps you already use daily. Each platform has a different setup process, so choose one to start with. You can always add more channels later.

Connecting to WhatsApp

WhatsApp is popular because most people already have it on their phones. OpenClaw integrates with WhatsApp through a QR code connection.

Setup Process:

During the onboarding wizard, when you reach the channel selection step:

1. Choose "WhatsApp" from the list

2. OpenClaw will display a QR code in your terminal

3. Open WhatsApp on your phone

4. Tap the three dots (Android) or Settings (iPhone)

5. Select "Linked Devices"

6. Tap "Link a Device"

7. Scan the QR code shown in your terminal

8. WhatsApp confirms the connection

How It Works: Once connected, you interact with OpenClaw by sending messages to yourself on WhatsApp. Yes, you're literally chatting with your own phone number. When you send a message to yourself, OpenClaw receives it, processes it through your AI model, and sends the response back.

This might feel weird at first, but it's actually convenient—you can access your AI assistant from any device where you have WhatsApp installed.

Important Security Note: OpenClaw uses WhatsApp Web protocol, which means your messages are end-to-end encrypted just like regular WhatsApp messages. However, remember that your AI provider (like Anthropic) can see the messages OpenClaw sends to their API.

Connecting to Telegram

Telegram is the easiest channel to set up because it has official bot support built into the platform. No QR codes or phone linking required. If you're looking for ways to expand your OpenClaw capabilities after installation, check out how to integrate OpenClaw with Telegram for seamless communication.

Setup Process:

1. Open Telegram and start a chat with @BotFather (this is an official Telegram bot that creates other bots)

2. Send the command /newbot

3. BotFather asks you to choose a name for your bot (like "My OpenClaw Assistant")

4. Then choose a username ending in "bot" (like "myclaw_helper_bot")

5. BotFather generates a token that looks like 1234567890:ABCdefGHIjklMNOpqrsTUVwxyz

6. Copy this token

7. In the OpenClaw onboarding wizard (or run openclaw config to configure later), select Telegram

8. Paste your bot token when prompted

That's it. Now you can open Telegram, search for your bot by username, and start chatting with your OpenClaw assistant.

Advantages of Telegram:

• Works on all your devices simultaneously

• Supports rich media (images, files, voice messages)

• No phone number required

• Fast and lightweight

Connecting to Discord

Discord integration is perfect if you want to share your AI assistant with a team or community server. For a comprehensive guide on building chat applications with OpenClaw, see how to create custom OpenClaw gateway chat apps.

Setup Process:

1. Go to discord.com/developers

2. Log in with your Discord account

3. Click "New Application" and give it a name

4. Navigate to the "Bot" section in the sidebar

5. Click "Add Bot"

6. Under the bot's username, click "Reset Token" to generate a new token

7. Copy the token (starts with something like MTAyNz...)

8. In OpenClaw, install the Discord skill:

openclaw skill install discord-bot

1. Configure the bot token:

openclaw config set discord.bot_token YOUR_BOT_TOKEN

1. Invite the bot to your Discord server using the OAuth2 URL generator in the developer portal

Once connected, you can mention your bot in any Discord channel where it has permissions, and it will respond with AI-generated answers.

Multi-Platform Setup

Here's the powerful part: you can connect all three (WhatsApp, Telegram, and Discord) to the same OpenClaw instance. They all share the same AI brain and memory.

This means you could:

• Check your calendar through WhatsApp while on your phone

• Have OpenClaw summarize a document via Telegram on your laptop

• Share AI-generated content with your team through Discord

All from one OpenClaw installation. To add additional channels after your initial setup, just run:

openclaw config

Select "Channels," then add whichever platforms you want.

How Do You Verify Your OpenClaw Installation Works?

After completing the installation and configuration, you should test that everything is working correctly before diving into advanced features.

Check Gateway Status

First, verify the OpenClaw gateway service is running:

openclaw status

You should see output indicating the gateway is active. If it shows as stopped or errored, try:

openclaw gateway restart

Access the Web Dashboard

Open your web browser and go to:

http://localhost:18789

You should see the OpenClaw dashboard interface. This is your control panel where you can:

• Chat with your AI assistant in the browser

• View active channels and their connection status

• Check logs and diagnostic information

• Manage skills and configurations

If the dashboard doesn't load, the gateway might not be running, or there could be a port conflict (another application using port 18789).

Send a Test Message Through Your Channel

The real test is sending a message through one of your connected channels.

For WhatsApp:

1. Open WhatsApp on your phone

2. Start a new chat and select your own number

3. Send a simple message like "Hello, are you working?"

4. Wait 5-10 seconds

5. You should receive an AI-generated response

For Telegram:

1. Open Telegram and find your bot

2. Send /start to activate the conversation

3. Send "What's the weather today?" or any other question

4. The bot should respond within a few seconds

For Discord:

1. Go to a channel where your bot has permissions

2. Mention the bot like @YourBotName hello

3. Wait for a response

If you get responses, congratulations! Your OpenClaw installation is working perfectly.

Run the Diagnostic Tool

OpenClaw includes a built-in health check tool that scans for common problems:

openclaw doctor

This command checks:

• Node.js version compatibility

• File permissions

• Configuration file validity

• API key connectivity

• Channel connection status

• Missing dependencies

If it finds problems, it suggests fixes. For automatic repairs, run:

openclaw doctor --fix

This command attempts to automatically resolve common issues like permission problems, missing directories, and configuration errors.

What Are Common OpenClaw Installation Errors and How Do You Fix Them?

Even with the automated installer, you might run into issues. Here are the most common problems and their solutions.

"openclaw: command not found"

Problem: After installation, running openclaw gives a "command not found" error.

Cause: The npm global bin directory isn't in your system PATH.

Solution:

Find where npm installs global packages:

npm config get prefix

Add that location to your PATH by editing your shell configuration file:

# For bash users

echo 'export PATH="$PATH:$(npm config get prefix)/bin"' >> ~/.bashrc

source ~/.bashrc

# For zsh users (default on modern macOS)

echo 'export PATH="$PATH:$(npm config get prefix)/bin"' >> ~/.zshrc

source ~/.zshrc

Try running openclaw again.

Node.js Version Errors

Problem: Installation fails with errors about Node.js version incompatibility.

Symptoms: Messages like "engine 'node' is incompatible" or crashes during startup.

Cause: You're running Node.js older than version 22.

Solution:

Update Node.js to version 22 or newer. On macOS:

brew upgrade node

On Linux:

curl -fsSL https://deb.nodesource.com/setup_22.x | sudo -E bash -

sudo apt-get install -y nodejs

Verify the version:

node --version

Then try installing OpenClaw again.

Permission Denied Errors

Problem: Installation fails with "permission denied" or "EACCES" errors.

Cause: npm is trying to write to system directories without proper permissions.

Wrong Solution: Don't run sudo npm install -g openclaw. This creates worse problems later.

Right Solution:

Configure npm to use a directory you own:

mkdir -p ~/.npm-global

npm config set prefix '~/.npm-global'

echo 'export PATH="$PATH:~/.npm-global/bin"' >> ~/.bashrc

source ~/.bashrc

Now reinstall OpenClaw without sudo:

npm install -g openclaw@latest

API Connection Failures

Problem: OpenClaw can't connect to your AI provider, showing errors like "unauthorized" or "invalid API key."

Symptoms: Messages fail to get responses, or you see authentication errors in logs.

Cause: API key is incorrect, expired, or not properly configured.

Solution:

1. Verify your API key is correct by logging into your AI provider's dashboard

2. Check for extra spaces or missing characters when you pasted it

3. Re-run the onboarding wizard with the correct key:

openclaw onboard

1. Make sure you have billing set up with your AI provider (most require payment info even for free tiers)

Gateway Won't Start or Keeps Crashing

Problem: The gateway starts but immediately crashes, or refuses to start at all.

Symptoms: openclaw status shows the gateway as stopped or errored.

Cause: Port conflict (another app using port 18789), corrupted config file, or leftover process from a previous installation.

Solution:

First, check if something else is using port 18789:

lsof -i :18789

If you see another process, stop it or configure OpenClaw to use a different port.

Check for old gateway processes:

ps aux | grep openclaw

Kill any stray processes:

pkill -f openclaw

Then restart the gateway:

openclaw gateway start

If problems persist, check the logs for specific error messages:

openclaw gateway logs

WhatsApp QR Code Won't Scan

Problem: The QR code appears in your terminal, but WhatsApp won't scan it or says it's invalid.

Cause: QR code expired (they're only valid for a few minutes), or terminal display issues.

Solution:

1. Make sure your terminal window is large enough to display the entire QR code

2. Try making your terminal font smaller to fit the whole code on screen

3. If the code is too distorted, restart the WhatsApp connection process:

openclaw config

Select WhatsApp channel and go through setup again. The wizard generates a fresh QR code.

1. Make sure you're scanning with the "Linked Devices" feature in WhatsApp, not the regular QR code scanner

Telegram Bot Not Responding

Problem: You message your Telegram bot, but it doesn't respond.

Cause: Incorrect bot token, bot not started, or gateway not running.

Solution:

1. Check gateway status:

openclaw status

If stopped, start it:

openclaw gateway start

1. Verify your bot token is correct in the configuration file

2. Make sure you sent /start to your bot before trying other messages

3. Check the gateway logs for error messages:

openclaw gateway logs

How Do You Install and Manage OpenClaw Skills?

Skills are what make OpenClaw truly powerful. They're like apps for your AI assistant—each skill teaches OpenClaw how to do something new. Understanding how skills work enhances your ability to build sophisticated automations. Learn more about creating your first OpenClaw skill.

What Are Skills?

Think of skills as instruction manuals for your AI. The base OpenClaw installation can chat and run basic commands, but skills teach it specialized capabilities:

• Email skill: Read, send, and organize emails

• Calendar skill: Manage appointments and schedules

• Browser skill: Search the web and extract information

• Code skill: Write, run, and debug code in multiple languages

• Smart home skill: Control lights, thermostats, and IoT devices

• Social media skills: Post updates, respond to messages

There are 53 official skills maintained by the OpenClaw team, plus hundreds more created by the community.

Finding Skills on ClawHub

ClawHub is the official registry for OpenClaw skills. It's like an app store for your AI assistant.

Visit https://clawhub.com in your browser to browse available skills. You can:

• Search by category (productivity, development, smart home, etc.)

• Read descriptions and documentation

• Check security ratings and community reviews

• See which skills are verified by the OpenClaw team

Important Security Warning: Some malicious skills have been caught trying to steal API keys or exfiltrate data. Always check the security rating and reviews before installing a skill, especially from unknown authors. ClawHub now partners with VirusTotal to provide automated security scanning for all published skills.

Installing Skills

There are three ways to install skills, from easiest to most advanced.

Method 1: Through the Web Dashboard (Easiest)

1. Open http://localhost:18789 in your browser

2. Navigate to the "Skills" section

3. Search for the skill you want

4. Click "Install"

5. Configure any required settings (like API keys for third-party services)

6. Enable the skill

Method 2: Using ClawHub CLI

For skills available on ClawHub:

npx clawdhub@latest install email-manager

Replace email-manager with whatever skill you want. The ClawHub CLI automatically downloads, installs, and configures the skill for you.

Method 3: Manual Installation

For custom or local skills:

1. Create a folder in ~/clawd/skills/ with your skill name

2. Add a SKILL.md file that describes what the skill does and when to use it

3. Include any additional scripts or configuration files

4. Restart the gateway:

openclaw gateway restart

Managing Installed Skills

List Your Installed Skills:

openclaw skills list

This shows all skills currently available to your assistant, with their status (enabled/disabled).

Disable a Skill:

If a skill is causing problems or you don't need it anymore, disable it without uninstalling:

openclaw skill disable skill-name

The skill stays installed but OpenClaw won't use it.

Enable a Disabled Skill:

openclaw skill enable skill-name

Update Skills:

Keep skills current to get bug fixes and new features:

openclaw skills update

This checks for updates to all installed skills and applies them.

Remove a Skill:

openclaw skill uninstall skill-name

Configuring Skills

Many skills need additional configuration like API keys for third-party services. For instance, if you want to add authentication to your skills, see how to implement authentication in OpenClaw skills.

Each skill's documentation explains its configuration requirements. Usually, you provide these settings through:

The Config File: Edit ~/.config/openclaw/openclaw.json5 and add skill-specific settings:

{

  skills: {

    "weather": {

      apiKey: "your-weather-api-key",

      defaultLocation: "San Francisco"

    }

  }

}

Environment Variables: Some skills read from environment variables:

export WEATHER_API_KEY="your-key-here"

Interactive Configuration: Run the skill's setup wizard (if it has one):

openclaw skill configure weather

Understanding Tools vs. Skills

This confuses many beginners, so let's clarify:

• Tools: Low-level capabilities built into OpenClaw (read files, run commands, make web requests, etc.)

• Skills: High-level instructions that teach the AI how to combine tools to accomplish complex tasks

For example:

• The "file read" tool lets OpenClaw read files

• The "bash execution" tool lets OpenClaw run commands

• The "code review" skill teaches OpenClaw how to use those tools together to analyze code quality

You don't install tools—they come with OpenClaw. You install skills to teach OpenClaw new ways to use those tools.

Frequently Asked Questions

How much does OpenClaw cost to run?

OpenClaw itself is completely free and open source. However, you pay for the AI model API usage. For typical personal use with Claude or GPT-4, expect $5-20 per month. Heavy usage can cost more. You control the budget by setting spending limits with your AI provider.

Is OpenClaw secure to use?

OpenClaw is experimental software with known security risks. Security researchers from Palo Alto Networks have raised concerns about its large codebase (430,000+ lines) and broad system access. Only install it on a dedicated machine or virtual environment—never on your primary computer with sensitive personal data. Think of it as powerful but not yet ready for production use.

Can I run OpenClaw on a Raspberry Pi?

Yes! OpenClaw works on Raspberry Pi 4 or newer with at least 4 GB of RAM. The Raspberry Pi Foundation even published an official guide. It's a great way to have a dedicated always-on AI assistant without running your main computer 24/7. However, performance will be slower than on a modern laptop.

What happens if I lose internet connection?

OpenClaw needs internet to communicate with AI model APIs and messaging platforms. If you lose connection, it can't process new requests. However, some skills with local-only functionality might still work. When connection restores, pending messages process automatically.

Can I use multiple AI models simultaneously?

Yes! You can configure OpenClaw to use different models for different tasks. For example, use GPT-4 for complex reasoning, Claude for writing tasks, and a cheaper model for simple questions. This is an advanced configuration that requires manual editing of the config file.

How do I uninstall OpenClaw?

To completely remove OpenClaw:

npm uninstall -g openclaw

rm -rf ~/.config/openclaw

rm -rf ~/clawd

Also disconnect any linked messaging channels (unlink WhatsApp devices, delete Telegram bots, etc.).

What's Next After Installation?

You've successfully installed OpenClaw, configured your AI model, and connected at least one messaging channel. Here are some great next steps:

1. Explore Official Skills: Install a few skills that match your daily workflow (email, calendar, notes)

2. Set Up Multiple Channels: Add more messaging platforms so you can access your assistant anywhere

3. Learn Advanced Commands: Read the OpenClaw documentation to discover powerful automation capabilities. For complex setups, check out how OpenClaw code agents handle local execution

4. Join the Community: Connect with other OpenClaw users on Discord or GitHub to share tips and troubleshooting help

5. Experiment with Automation: Try having OpenClaw automate repetitive tasks like organizing files or summarizing daily news

OpenClaw is a powerful tool that gets more useful the more you explore its capabilities. Start simple, experiment often, and gradually build up to more complex automations as you get comfortable.

The most important thing is to remember that this is experimental software. Always verify important information it provides, and never rely on it for critical tasks without human oversight. Enjoy your new AI assistant!