OpenClaw Center

Install OpenClaw

Choose your installation method. OpenClaw runs on macOS, Linux, and Windows.

Choose Your Method

Install Script

Fastest way to get started. One command installs everything.

curl -fsSL ... | bash

Docker

Run in a container. Best for servers and isolated environments.

docker run ...

From Source

Clone and run directly. Best for developers who want to contribute.

git clone ...

Platform-Specific Guides

🍎

macOS

Install on Mac using Homebrew or the install script. Works great on Mac mini, Mac Studio, or any Mac.

macOS Setup →Mac mini Guide →

🪟

Windows

Use WSL2 or PowerShell installer. Linux via WSL2 recommended for best experience.

Windows Setup Guide →

🐧

Linux

Install on Ubuntu, Debian, Fedora, or any Linux distro. Works great on Raspberry Pi.

Ubuntu Guide →Other Linux →

🏠

Home Lab / Server

Run on a home server or cloud VPS. Docker installation recommended.

Docker Guide →

System Requirements

Hardware

• 4 GB RAM minimum
• 8 GB recommended
• 64-bit processor

Software

• Node.js 22+
• macOS / Linux / Windows
• Docker (optional)

Model Backend

• Ollama (local)
• OpenAI API
• Anthropic API

View full requirements →

Docker vs Local Install

🐳 Docker Install

• Best for servers & VPS
• Isolated from system
• Easy to update
• Requires Docker installed
• Port forwarding needed

Docker Guide →

💻 Local Install

• Best for Mac/Desktop
• Direct access to files
• No container overhead
• Requires Node.js 22+
• Easier debugging

macOS Guide →

macOS vs Linux

Feature	macOS	Linux
Install Script	✅ Works	✅ Works
Homebrew	✅ Supported	❌ Not available
Raspberry Pi	❌ Not supported	✅ Supported
Systemd Service	✅ LaunchAgent	✅ Systemd
Best For	Mac mini, Desktop	Server, VPS, Pi

Minimum Hardware Specs

☁️ Cloud API

Using OpenAI, Anthropic, or other cloud APIs:

• 4 GB RAM minimum
• Any modern CPU
• Internet connection

🧠 Local (CPU)

Running Ollama with CPU-only:

• 8 GB RAM minimum
• Multi-core CPU
• 10GB+ disk space

🎮 Local (GPU)

Running Ollama with GPU (recommended):

• 16 GB VRAM minimum
• NVIDIA GPU
• CUDA drivers

Recommended Hardware by Use Case

🍎 OpenClaw on Mac mini

Mac mini is perfect for always-on OpenClaw deployment. M1/M2/M4 chips handle local models well.

• Recommended: Mac mini M2 Pro or newer (16GB RAM)
• Minimum: Any Mac mini with 8GB RAM
• Storage: 256GB+ SSD recommended

🏠 Home Lab (Raspberry Pi)

Run OpenClaw on a Raspberry Pi 4 or 5 for a low-power home server.

• Recommended: Raspberry Pi 5 with 8GB RAM
• Storage: 64GB+ microSD or SSD via USB
• Use: Cloud APIs (not local models)

🖥️ Desktop Linux

Full Linux desktop or laptop installation.

• Recommended: 16GB RAM, modern CPU
• Minimum: 8GB RAM, any 64-bit CPU
• GPU: Optional for local models

What's Next?

Quick Start

Get OpenClaw running in 5 minutes with the step-by-step guide.

Choose a Model

Learn about compatible AI models and how to configure them.

Troubleshooting

Common issues and how to fix them.

Next Steps

Ready to configure OpenClaw? The Setup Guide walks you through:

• Running the configuration wizard
• Connecting chat channels (Telegram, Discord, WhatsApp)
• Setting up your AI model
• Starting the gateway and testing
• Optional: reverse proxy, SSL, automation

OpenClaw Setup Guide →

Need Help?

Join our community to get help with installation.