What is OpenClaw?

OpenClaw is an open-source personal AI assistant that runs on your own hardware. Originally released as Clawdbot in November 2025 by Peter Steinberger (founder of PSPDFKit), the project exploded in popularity—gaining over 68,000 GitHub stars within its first week.

After a trademark request from Anthropic (the name "Clawd" sounded too similar to "Claude"), the project briefly became Moltbot before settling on its current name: OpenClaw. The lobster mascot stuck around though, and you'll see references to "molting" and "the lobster way" throughout the community.

How Does It Work?

At its core, OpenClaw acts as a local gateway that connects AI language models to your machine and your messaging apps. Here's the basic flow:

1. You send a message via WhatsApp, Telegram, Discord, or the web dashboard
2. OpenClaw receives it through the Gateway process running on your server
3. The message goes to your LLM (Claude, GPT-4, or a local model like Llama)
4. The AI responds and can optionally execute actions (run scripts, read files, browse the web)
5. You get the response back in your chat app

The key difference from services like ChatGPT is that everything runs on hardware you control. Your conversations stay private. Your API keys never leave your machine.

Why Use ClawBook Instead of Self-Hosting?

Setting up OpenClaw yourself isn't trivial. You need to:

• Provision a server (or dedicate a home machine that's always on)
• Install Node.js 22+, Docker, and various dependencies
• Configure firewalls and security settings
• Set up SSL certificates for secure connections
• Handle WhatsApp Web pairing (which can be finicky)
• Keep everything updated and maintained

ClawBook handles all of this for you. We provide pre-configured VPS instances with OpenClaw ready to go. You just connect your LLM provider, scan a QR code for WhatsApp, and you're done.

DIY Self-Hosting	ClawBook Managed
2-4 hours setup	5-minute setup
You manage updates	We handle updates
You fix issues	We provide support
Variable costs	Predictable pricing
Your hardware	Enterprise-grade VPS

What Can OpenClaw Actually Do?

Once connected, your AI assistant can:

• Chat naturally across WhatsApp, Telegram, Discord, Slack, iMessage, and more
• Remember context with persistent memory stored as local Markdown files
• Execute shell commands (within a Docker sandbox for safety)
• Read and write files in your designated workspace
• Browse the web to research information or fill out forms
• Connect to 50+ integrations including smart home devices and productivity tools

The model-agnostic design means you can use whichever LLM fits your needs. Most users connect Anthropic's Claude (the project is optimized for it), but OpenAI, Google, and fully local models via Ollama all work.

ClawBook Architecture

When you sign up for ClawBook, here's what runs on your dedicated VPS:

┌─────────────────────────────────────────────────────────┐
│                    Your ClawBook VPS                    │
├─────────────────────────────────────────────────────────┤
│                                                         │
│  ┌───────────────┐     ┌───────────────┐              │
│  │   OpenClaw    │     │    Docker     │              │
│  │   Gateway     │────▶│   Sandbox     │              │
│  │  (port 18789) │     │  (isolation)  │              │
│  └───────┬───────┘     └───────────────┘              │
│          │                                             │
│  ┌───────┴───────────────────────────────────┐        │
│  │            Channel Bridges                 │        │
│  ├──────────┬──────────┬──────────┬─────────┤        │
│  │ WhatsApp │ Telegram │ Discord  │ Slack   │        │
│  │ (Baileys)│ (grammY) │          │         │        │
│  └──────────┴──────────┴──────────┴─────────┘        │
│                                                         │
└─────────────────────────────────────────────────────────┘
                          │
                          ▼
              ┌───────────────────────┐
              │   Your LLM Provider   │
              │ (Anthropic / OpenAI / │
              │   Local via Ollama)   │
              └───────────────────────┘

The Gateway is the heart of the system—a single long-running process that maintains connections to all your messaging channels and routes messages to your AI model.

Getting Started

Ready to set up your ClawBook instance? Here's the path:

1. Quick Start Guide — Get connected in 5 minutes
2. Connect Your LLM — Set up Anthropic, OpenAI, or Ollama
3. WhatsApp Setup — Link your WhatsApp account
4. Security Hardening — Lock down your instance

Getting Help

• Email Support: support@clawbook.io
• Documentation: You're reading it
• Official OpenClaw Docs: docs.openclaw.ai
• GitHub Issues: github.com/openclaw/openclaw

ClawBook is an independent hosting provider—we're not affiliated with the OpenClaw project or Anthropic. We just love the technology and want to make it accessible to everyone.