Overview

Advantages

  • Simplest setup
  • No network exposure
  • Full control
  • No hosting costs

Considerations

  • Must be running locally
  • No remote access by default
  • Depends on your machine

Requirements

  • Node.js 22+: Node.js version 22 or higher is required
  • macOS, Linux, or WSL2: Supported operating systems
  • npm or pnpm: Package manager for installation

Setup Steps

  1. Install OpenClaw: Install OpenClaw globally via npm or the installer script 
    curl -fsSL https://openclaw.bot/install.sh | bash
  2. Run Onboarding: Complete the setup wizard 
    openclaw onboard --install-daemon
  3. Configure Providers: Add your AI provider credentials during onboarding or manually
  4. Start Gateway: Start the Gateway service 
    openclaw gateway start
  5. Verify Status: Check that everything is running correctly 
    openclaw status

Configuration Example

{
  "gateway": {
    "mode": "local",
    "bind": "loopback",
    "port": 18789,
    "auth": {
      "mode": "token",
      "allowTailscale": false
    }
  }
}

Deploy with AI Providers

Choose your AI provider for local deployment deployment:

Anthropic

Creator of Claude AI models with advanced reasoning capabilities

OpenAI Codex

GPT models with broad capabilities and wide ecosystem support

Google Gemini

Multimodal AI with competitive pricing and speed

OpenRouter

Unified API gateway to multiple AI providers

Ollama

Run open-source LLMs locally with no API costs

AWS Bedrock

Enterprise AI through AWS with compliance and security

OpenAI (Standard)

Standard OpenAI API for GPT models

Frequently Asked Questions

What are the requirements for Local Deployment?
Local Deployment requires: Node.js 22+, macOS, Linux, or WSL2, npm or pnpm. The setup typically takes 15-30 minutes for first-time users.
Where is state stored with Local Deployment?
State is stored in ~/.openclaw/. This includes configuration, credentials, session history, and workspace files. Always back up this directory.