OpenClaw: Install, Set Up, and Start Making Money With Your Own AI Agent (2026)

Disclosure: This post contains affiliate links. If you purchase through these links, I may earn a commission at no extra cost to you. I only recommend products I genuinely believe in.

There’s a new kind of tool blowing up in the AI space, and it’s not another chatbot wrapper. It’s OpenClaw — an open-source AI agent framework with over 240,000 stars on GitHub that lets you run your own personal AI assistant on your own devices.

Unlike ChatGPT or Claude’s web interfaces, OpenClaw doesn’t live in a browser tab. It runs as a service on your machine (or a server), connects to your messaging apps — WhatsApp, Telegram, Discord, Slack, and more — and can execute real tasks: send messages, run scripts, manage files, control smart home devices, automate workflows, and interact with any API you point it at.

Think of it as your own AI employee that works 24/7, responds to your messages, and handles whatever you train it to do.

This guide covers how to install it, set it up, and — most importantly — how people are actually using it to make money in 2026.

What Is OpenClaw?

OpenClaw is a self-hosted AI agent runtime. Here’s what that means in plain English:

• Self-hosted — it runs on your computer or server, not someone else’s cloud
• AI agent — it doesn’t just answer questions; it takes actions (sends messages, runs code, manages files, calls APIs)
• Runtime — it’s always running in the background, waiting for your instructions

The architecture is straightforward: a Gateway (WebSocket server) sits at the center, connecting your chat channels (WhatsApp, Telegram, etc.) to an AI model (Claude, GPT-4, local models) that has access to tools and skills (shell commands, file operations, web browsing, custom scripts).

You message OpenClaw from your phone. It thinks, executes, and replies.

Key Features

• 15+ chat channels: WhatsApp, Telegram, Discord, Slack, Signal, iMessage, Microsoft Teams, Matrix, and more
• Any AI model: Claude, GPT-4/GPT-4o, Gemini, Mistral, Llama (local), or any OpenAI-compatible API
• 50+ built-in skills: Email, notes, weather, Spotify control, image generation, PDF editing, voice calls, smart home, and more
• Custom skills: Build your own automation skills with scripts and templates
• Cron scheduling: Run tasks on a schedule (daily reports, content generation, monitoring)
• Multi-agent routing: Run multiple isolated agents for different purposes
• Browser automation: Control a dedicated Chrome instance for web scraping and testing
• Voice mode: Speech recognition and text-to-speech with ElevenLabs
• 100% open-source: No vendor lock-in, full transparency, active community

Installation: Step by Step

Requirements

• Node.js 22 or later (check with node --version)
• macOS, Linux, or Windows (WSL2)
• A terminal you’re comfortable with

Step 1: Install OpenClaw

npm install -g openclaw@latest

This installs the openclaw CLI globally. Verify it worked:

openclaw --version

Step 2: Run the Onboarding Wizard

openclaw onboard --install-daemon

The wizard walks you through:

1. Gateway setup — starts the local WebSocket server that everything connects to
2. AI model configuration — connect your API key for Claude, OpenAI, or another provider
3. Channel setup — link your WhatsApp, Telegram, or other messaging apps
4. Workspace configuration — set the directory where OpenClaw stores its data and skills
5. Daemon installation — sets OpenClaw to start automatically on boot

Step 3: Connect WhatsApp (or Your Preferred Channel)

openclaw channels login --verbose

For WhatsApp, this shows a QR code in your terminal. Scan it with the WhatsApp app on your phone (Settings > Linked Devices > Link a Device). Once connected, any message you send to yourself on WhatsApp goes to OpenClaw.

Step 4: Verify Everything Works

openclaw doctor

This runs health checks on your gateway, channels, and model connections. Fix any issues it flags before moving on.

Step 5: Send Your First Message

Open WhatsApp (or your connected channel) and send a message. OpenClaw should respond using the AI model you configured.

That’s it. You have a personal AI agent running on your machine, accessible from your phone.

Key Commands You’ll Use Daily

# Check status of all channels
openclaw status

# View logs in real-time
openclaw logs

# Manage cron jobs (scheduled tasks)
openclaw cron list
openclaw cron add "daily-report" --schedule "0 9 * * *" --command "..."

# List available skills
openclaw skills list

# Send a message from the terminal
openclaw message send --target +1234567890 --message "Hello from OpenClaw"

# Run the agent on a specific task
openclaw agent --message "Summarize my unread emails" --deliver

# Open the web dashboard
openclaw dashboard

# Health check
openclaw doctor

Setting Up for 24/7 Operation

Running OpenClaw on your laptop works for testing, but to make money with it, you need it running around the clock. Two options:

Option A: Run on a VPS ($5-10/month)

A cheap VPS from DigitalOcean, Hetzner, or any provider works. Requirements are minimal: 2GB RAM, 1 CPU core, 20GB storage.

# On your VPS:
npm install -g openclaw@latest
openclaw onboard --install-daemon

The daemon flag installs it as a system service that starts on boot and restarts on crash.

Option B: Run on a Spare Machine at Home

An old laptop, a Raspberry Pi 4/5, or a Mac Mini works perfectly. Just install, run the onboard wizard, and leave it running. Use Tailscale (free) for secure remote access.

5 Ways People Are Making Money With OpenClaw in 2026

Here’s where it gets interesting. OpenClaw isn’t just a cool tech project — people are building real income streams with it.

1. AI-Powered Customer Service Bot ($500-$3,000/month per client)

The setup: Deploy an OpenClaw instance connected to a client’s WhatsApp Business or Telegram channel. Load it with their FAQ, product catalog, and standard operating procedures via custom skills.

What it does: Handles customer inquiries 24/7 — answers questions, provides order status, collects leads, and escalates to a human only when necessary.

Why it works: Small businesses are desperate for this. They can’t afford a customer service team, but they lose sales when nobody responds to messages at 10 PM. You charge $500-$3,000/month per client depending on complexity.

Your cost: One VPS ($5-10/month) can run multiple isolated agents for different clients. AI API costs typically run $20-50/month per client depending on volume.

2. Content Automation Agency ($1,000-$5,000/month)

The setup: Build an OpenClaw skill that generates, reviews, and publishes content on a schedule. Connect it to your client’s CMS, social media, and email platform.

What it does: Researches keywords, generates articles, optimizes for SEO, schedules social media posts, and sends email newsletters — all automated with human approval via WhatsApp.

Why it works: Content marketing is expensive. A freelance writer charges $200-500 per article. An OpenClaw content pipeline produces first drafts at a fraction of the cost, with your editorial oversight maintaining quality.

This is exactly how we run bestaistack.net — OpenClaw generates drafts, sends them to WhatsApp for approval, and publishes to the site. The entire pipeline is automated.

3. AI Automation Consulting ($2,000-$10,000 per project)

The setup: You become the expert who sets up OpenClaw for businesses. Install it, configure their channels, build custom skills for their specific workflows, and train them on basic usage.

What it does: You’re selling your expertise, not a product. Each client gets a custom agent tailored to their business — from appointment scheduling to inventory alerts to automated reporting.

Why it works: Most business owners have heard of AI but have no idea how to implement it. You bridge that gap. A single setup project can pay $2,000-$10,000 depending on complexity, with recurring monthly maintenance fees of $200-500.

4. Lead Generation Bot ($500-$2,000/month per client)

The setup: Create an OpenClaw agent that monitors specified channels (web chat, WhatsApp, Telegram), qualifies incoming leads by asking qualifying questions, enriches lead data, and pushes qualified leads to the client’s CRM.

What it does: Automates the top of the sales funnel. Instead of a sales rep spending time on tire-kickers, the AI qualifies leads 24/7 and only escalates warm prospects.

Why it works: Lead qualification is tedious and expensive. Businesses pay $15-25/hour for SDRs who can only work 8 hours. An OpenClaw agent costs a fraction of that and works around the clock.

5. Personal AI Assistant as a Service ($200-$500/month per client)

The setup: Deploy personalized OpenClaw instances for busy professionals — executives, founders, creators. Configure it with their calendar, email, notes, and preferences.

What it does: Acts as a virtual executive assistant — manages scheduling, drafts emails, summarizes documents, sets reminders, researches topics on demand, all accessible via the client’s preferred messaging app.

Why it works: A human virtual assistant costs $2,000-5,000/month. An AI-powered one costs you $30-50/month in API and hosting, and you charge $200-500. The margins are excellent.

Quick Profitability Calculator

The economics are compelling because your costs are almost entirely AI API usage and hosting — both of which are cheap and scale linearly. There’s no inventory, no physical product, and minimal overhead.

Pairing OpenClaw With the Right Tools

OpenClaw handles the AI agent layer, but a complete business needs more:

• For funnels, email, and course hosting: Systeme.io gives you everything on a free plan. Connect OpenClaw to its API for automated email triggers and funnel updates. See our full Systeme.io review.

• For choosing the right AI model: OpenClaw works with multiple models. Our ChatGPT vs Claude vs Gemini comparison helps you pick the best one for your use case.

• For workflow automation beyond OpenClaw: Combine it with n8n, Make, or Zapier for complex multi-app workflows.

• For AI-powered copywriting: Use Jasper or Copy.ai alongside OpenClaw for specialized marketing content.

Common Setup Issues and Fixes

“Gateway won’t start” — Usually a port conflict. Run openclaw gateway --force to kill anything on the default port, or specify a custom port with --port 18789.

“WhatsApp QR code expired” — QR codes expire quickly. Run openclaw channels login again and scan within 30 seconds.

“Model not responding” — Check your API key configuration with openclaw config get. Verify the model provider is accessible with openclaw models list.

“Skills show as ‘missing’” — Some skills require external CLIs. Run openclaw doctor to see which dependencies are needed and install them.

“High API costs” — Use local models (Llama via Ollama) for simple tasks and reserve Claude/GPT-4 for complex reasoning. Configure model fallbacks in your config to optimize costs:

openclaw config set agents.defaults.model.primary "claude-cli/sonnet"
openclaw config set agents.defaults.model.fallbacks '["ollama/llama3.1:latest"]'

Getting Started Today

1. Install OpenClaw — npm install -g openclaw@latest && openclaw onboard --install-daemon
2. Connect WhatsApp — openclaw channels login
3. Test it — Send yourself a message and watch it respond
4. Pick a business model from the 5 above and start building your first skill
5. Set up a VPS for 24/7 operation when you’re ready to serve clients

The barrier to entry is lower than most people think. If you can install an npm package and follow a setup wizard, you can run your own AI agent business. The real skill isn’t the technology — it’s understanding what businesses need and packaging your agent to solve those problems.

For a broader view of making money with AI tools, check out our complete beginner’s guide.

Last updated: February 2026. OpenClaw is actively developed — check the official docs and GitHub for the latest features.