How to Install OpenClaw: Step-by-Step Setup Guide (10 Minutes)

TL;DR

Setting up OpenClaw takes 10-15 minutes. Install Node.js, run the OpenClaw CLI wizard, connect WhatsApp (or Telegram/Discord), configure your LLM provider, and send your first message. This guide covers every step with troubleshooting tips for Mac, Windows, and Linux, plus how to integrate with Apidog for API development workflows.

Prerequisites

Before you start, make sure you have:

Node.js 18 or higher - Download from nodejs.org
A messaging platform account - WhatsApp, Telegram, or Discord
An LLM API key - Claude (Anthropic), OpenAI, or a local model
Terminal access - Command Prompt (Windows), Terminal (Mac/Linux)
10-15 minutes - For the complete setup
Optional: Apidog account - For API development automation

button

System Requirements

Component	Minimum	Recommended
RAM	2GB	4GB+
Storage	500MB	1GB+
OS	macOS 10.15+, Windows 10+, Linux	Latest versions
Internet	Required for cloud LLMs	Stable connection

Step 1: Install Node.js

OpenClaw runs on Node.js. Check if you already have it:

node --version

If you see v18.0.0 or higher, skip to Step 2. Otherwise, install Node.js:

macOS

Option A: Official Installer

Visit nodejs.org
Download the LTS version
Run the installer
Verify: node --version

Option B: Homebrew

brew install node

Windows

Visit nodejs.org
Download the Windows installer (.msi)
Run the installer (accept defaults)
Restart your terminal
Verify: node --version

Linux (Ubuntu/Debian)

curl -fsSL https://deb.nodesource.com/setup_20.x | sudo -E bash -
sudo apt-get install -y nodejs

Verify installation:

node --version
npm --version

Step 2: Install OpenClaw CLI

Open your terminal and run:

npm install -g openclaw

This installs the OpenClaw command-line tool globally. The process takes 1-2 minutes.

Troubleshooting:

Error: Permission denied (Mac/Linux)

sudo npm install -g openclaw

Error: npm not found (Windows)

Restart your terminal after installing Node.js
Add Node.js to your PATH manually if needed

Verify installation:

openclaw --version

You should see the version number (e.g., 2.1.0).

Step 3: Run the Onboarding Wizard

Start the setup wizard:

openclaw init

The wizard guides you through configuration. Answer each prompt:

3.1: Choose Your LLM Provider

? Which LLM provider do you want to use?
  > Anthropic (Claude)
    OpenAI (GPT-4)
    Local Model (Ollama)
    Custom API

Recommendation: Claude Sonnet 4.6 offers the best balance of speed, quality, and cost.

3.2: Enter Your API Key

? Enter your Anthropic API key:

Get your API key:

Anthropic: console.anthropic.com
OpenAI: platform.openai.com

Paste the key and press Enter. The wizard validates it automatically.

3.3: Select Messaging Platform

? Which messaging platform do you want to connect?
  > WhatsApp (recommended)
    Telegram
    Discord
    Slack
    iMessage (macOS only)

WhatsApp is the most popular choice. It works on all platforms and requires no bot setup.

3.4: Configure Storage

? Where should OpenClaw store data?
  > ~/.openclaw (default)
    Custom path

Accept the default unless you have a specific reason to change it.

3.5: Enable Tools

? Which tools should OpenClaw have access to?
  [x] read - Read files
  [x] write - Write files
  [x] exec - Execute commands
  [x] web_search - Search the internet
  [x] web_fetch - Fetch web pages
  [x] memory - Store memories
  [x] schedule - Schedule tasks

Security Note: Only enable tools you trust. Start with read, write, web_search, and memory. Add exec later if needed.

Step 4: Connect WhatsApp

After the wizard completes, OpenClaw displays a QR code in your terminal.

4.1: Scan the QR Code

Open WhatsApp on your phone
Tap the three dots (Android) or Settings (iPhone)
Select “Linked Devices”
Tap “Link a Device”
Scan the QR code in your terminal

4.2: Verify Connection

Once scanned, you’ll see:

✓ WhatsApp connected successfully
✓ OpenClaw is ready to use

OpenClaw creates a new chat in your WhatsApp. Send a message to test it.

Troubleshooting WhatsApp Connection

QR code doesn’t appear:

Check your internet connection
Restart the setup: openclaw init --reset
Try a different terminal (some don’t render QR codes well)

QR code expires:

The code is valid for 60 seconds
Run openclaw connect whatsapp to generate a new one

Connection fails after scanning:

Make sure WhatsApp is updated to the latest version
Check firewall settings (OpenClaw needs port 3000)
Restart your phone and try again

Step 5: Alternative Platforms

Telegram Setup

Run: openclaw connect telegram
Visit t.me/BotFather
Send /newbot and follow instructions
Copy the bot token
Paste it into OpenClaw
Start chatting with your bot

Discord Setup

Run: openclaw connect discord
Visit discord.com/developers
Create a new application
Go to “Bot” section and create a bot
Copy the bot token
Paste it into OpenClaw
Invite the bot to your server

Slack Setup

Run: openclaw connect slack
Visit api.slack.com/apps
Create a new app
Enable Socket Mode
Copy the app token
Paste it into OpenClaw
Install the app to your workspace

Step 6: Start OpenClaw

Run the OpenClaw agent:

openclaw start

You’ll see:

🦞 OpenClaw is running
📱 Connected to WhatsApp
🤖 Using Claude Sonnet 4.6
💾 Data stored in ~/.openclaw

Keep this terminal window open. OpenClaw runs in the foreground.

Run in Background (Optional):

openclaw start --daemon

This runs OpenClaw as a background process. Check status with:

openclaw status

Stop it with:

openclaw stop

Step 7: Send Your First Message

Open WhatsApp and find the OpenClaw chat. Send a message:

You: “Hello”

OpenClaw: “Hi! I’m OpenClaw, your AI assistant. I can help you with files, web searches, scheduling tasks, and more. What would you like to do?”

You: “What’s the weather in San Francisco?”

OpenClaw: [Searches the web and returns current weather]

You: “Create a file called test.txt with ‘Hello World’ in it”

OpenClaw: “Done. Created test.txt in your current directory.”

Verify the file exists:

cat test.txt

You should see: Hello World

Step 8: Integrate with Apidog (For API Developers)

If you’re an API developer, integrate OpenClaw with Apidog to automate your API workflow.

8.1: Install Apidog CLI

npm install -g apidog-cli

8.2: Configure Apidog Integration

# Set your Apidog API key
openclaw config set integrations.apidog.apiKey "YOUR_APIDOG_API_KEY"

# Set your project ID
openclaw config set integrations.apidog.projectId "YOUR_PROJECT_ID"

# Enable Apidog integration
openclaw config set integrations.apidog.enabled true

Get your Apidog API key from Apidog Settings.

8.3: Test Apidog Integration

Send a message to OpenClaw:

You: “Run Apidog tests for my API”

OpenClaw: "Running Apidog test suite…

✓ Authentication tests: 12/12 passed ✓ User endpoints: 8/8 passed ✓ Payment API: 15/15 passed

All tests passed!

8.4: Set Up API Monitoring

You: “Monitor my production API using Apidog. Check every 5 minutes.”

OpenClaw: "Monitoring enabled. I’ll check your API every 5 minutes using Apidog monitors and alert you if anything fails.

Monitored endpoints:

POST /api/auth/login
GET /api/users
POST /api/payments
GET /api/products

8.5: Automate Documentation Updates

You: “When I update API endpoints in Apidog, notify me and update our docs”

OpenClaw: "Automation configured. I’ll watch for Apidog project changes and:

Notify you when endpoints are added/modified
Update documentation automatically
Generate changelog entries

Configuration Tips

Customize Your Experience

Edit the config file:

openclaw config edit

This opens ~/.openclaw/config.json in your default editor.

Useful Settings

Change the LLM model:

{
  "llm": {
    "provider": "anthropic",
    "model": "claude-sonnet-4-6",
    "temperature": 0.7
  }
}

Adjust memory settings:

{
  "memory": {
    "maxMessages": 100,
    "persistAcrossSessions": true
  }
}

Set default working directory:

{
  "workspace": {
    "defaultPath": "/Users/yourname/projects"
  }
}

Configure Apidog integration:

{
  "integrations": {
    "apidog": {
      "enabled": true,
      "apiKey": "your_api_key",
      "projectId": "your_project_id",
      "autoTest": true,
      "autoMonitor": true,
      "autoDoc": true
    }
  }
}

Save the file and restart OpenClaw:

openclaw restart

Common Setup Issues

Issue 1: “Command not found: openclaw”

Cause: npm global bin directory not in PATH

Fix (Mac/Linux):

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

Fix (Windows):

Search for “Environment Variables”
Edit PATH
Add: C:\Users\YourName\AppData\Roaming\npm
Restart terminal

Issue 2: “API key invalid”

Cause: Incorrect or expired API key

Fix:

Get a new key from your LLM provider
Run: openclaw config set llm.apiKey YOUR_NEW_KEY
Restart OpenClaw

Issue 3: “Port 3000 already in use”

Cause: Another application is using port 3000

Fix:

openclaw start --port 3001

Or kill the process using port 3000:

Mac/Linux:

lsof -ti:3000 | xargs kill -9

Windows:

netstat -ano | findstr :3000
taskkill /PID <PID> /F

Issue 4: “WhatsApp disconnected”

Cause: Phone went offline or unlinked the device

Fix:

Run: openclaw connect whatsapp
Scan the new QR code
Keep your phone connected to the internet

Issue 5: High CPU usage

Cause: OpenClaw is processing a long-running task

Fix:

Check what OpenClaw is doing: Send “What are you working on?”
Stop background tasks: openclaw tasks clear
Reduce heartbeat frequency in config

Platform-Specific Notes

macOS

iMessage integration requires macOS 12+ and Messages.app access
Grant terminal permissions in System Preferences > Privacy & Security
Use Homebrew for easier Node.js management

Windows

Run terminal as Administrator for global npm installs
Windows Defender may flag OpenClaw (add exception if needed)
Use Windows Terminal for better QR code rendering

Linux

Some distros require build-essential for npm packages:
Run OpenClaw as a systemd service for auto-start:

sudo apt-get install build-essential

openclaw install-service

Running OpenClaw 24/7

Option 1: Raspberry Pi

Many users run OpenClaw on a Raspberry Pi 4 or 5:

Install Raspberry Pi OS
Follow the standard setup steps
Enable auto-start:

openclaw install-service

Cost: $50-100 for hardware, $0/month for hosting

Option 2: Cloud Server

Run OpenClaw on a VPS (DigitalOcean, Linode, AWS):

Create a small instance ($5-10/month)
SSH into the server
Install Node.js and OpenClaw
Run as a background service

Cost: $5-10/month

Option 3: Always-On Computer

If you have a computer that’s always on (home server, old laptop):

Install OpenClaw
Set it to start on boot
Keep the machine running

Cost: $0/month (electricity only)

Next Steps

Now that OpenClaw is running, explore what it can do:

Install skills - Add pre-built workflows for common tasks
Configure tools - Enable or disable capabilities
Set up automation - Create scheduled tasks with Heartbeat
Join the community - Get help and share your setup
Integrate with Apidog - Automate your API development workflow with Apidog

Try these commands:

“List available skills”
“Show me what tools I have enabled”
“Schedule a reminder for tomorrow at 9 AM”
“Search for the latest AI news”
“Run my Apidog test suite” (if Apidog is configured)

Quick Start: OpenClaw + Apidog Workflow

If you’re an API developer, here’s a quick workflow to get started:

1. Set Up API Monitoring

You: "Monitor my API endpoints using Apidog. Alert me if any fail."

OpenClaw: "Monitoring configured. Checking every 5 minutes."

2. Automate Testing After Deployments

You: "After each deployment, run Apidog tests and send me results"

OpenClaw: "Automation configured. I'll run tests after deployments."

3. Keep Documentation Updated

You: "When API endpoints change in Apidog, update our documentation"

OpenClaw: "Documentation automation enabled."

4. Track API Performance

You: "Send me weekly API performance reports from Apidog"

OpenClaw: "Scheduled. You'll receive reports every Monday at 9 AM."

FAQ

Q: Do I need to keep my terminal open?A: Yes, unless you run OpenClaw as a daemon (openclaw start --daemon).

Q: Can I use multiple messaging platforms?A: Yes, connect as many as you want: openclaw connect telegram, openclaw connect discord, etc.

Q: How do I update OpenClaw?A: Run npm update -g openclaw and restart.

Q: Can I use a local LLM instead of API keys?A: Yes, install Ollama and configure OpenClaw to use it. No API costs.

Q: Is my data encrypted?A: OpenClaw stores data locally in plain text. Encrypt your disk for full protection.

Q: Can I run OpenClaw on Android/iOS?A: Not directly. Run it on a computer or server and connect via messaging apps.

Q: How much do API calls cost?A: Depends on usage. Typical users spend $5-20/month. Monitor costs in your LLM provider dashboard.

Q: Can I share OpenClaw with my team?A: Yes, but each person needs their own instance. Enterprise features are in development.

Q: Does Apidog integration cost extra?A: No, Apidog integration is free. You need an Apidog account (free tier available).

Q: Can OpenClaw work without Apidog?A: Yes, Apidog integration is optional. OpenClaw works standalone for general automation.

Conclusion

You now have OpenClaw running and connected to your messaging platform. The setup takes 10-15 minutes, but the productivity gains last forever.

Start with simple tasks—file operations, web searches, reminders—and gradually explore advanced features like skills, automation, and custom tools.

For API developers, integrating with Apidog unlocks powerful workflows: automated testing, endpoint monitoring, documentation updates, and performance tracking—all through simple text messages.

If you run into issues, check the troubleshooting section or ask the OpenClaw community on Discord.

button