Frequently Asked Questions

Yes! openclaw.ai is completely free and open-source under the MIT license. You can use it for personal, commercial, or enterprise projects without any licensing fees. However, you may incur costs from third-party API providers (like OpenAI or Anthropic) if you choose to use their services.

Core requirements:

• Node.js 22.22.3+, 24.15+, or 25.9+ (Node 26 recommended)
• 64-bit operating system
• Internet connection (for cloud LLM APIs)

Resource use varies with channels, browser automation, plugins, and model traffic. Docker image builds need at least 2GB RAM; local models need substantially more memory and storage based on the model.

The official installer supports the major desktop and server platforms:

• Windows: native Windows Hub, PowerShell CLI installer, or WSL2 Gateway
• macOS: Intel and Apple Silicon
• Linux: common 64-bit distributions with Node support
• Containers: Docker, Kubernetes
• ARM64: supported where Node and required dependencies are available

The installer itself is usually quick. Total setup time depends on downloading Node, provider authentication, channel pairing, network speed, and optional plugins. Skip optional onboarding steps and return later with openclaw configure.

Our beginner tutorials require no prior programming experience. You'll be copying and pasting commands with clear explanations of what each step does. However, for advanced tutorials (production deployment, Kubernetes), basic familiarity with command-line interfaces and server administration is helpful.

Yes! You can install openclaw.ai using the install script which handles permissions automatically. Alternatively, you can configure npm to install global packages in your home directory. However, some features (like running the daemon on startup) may require administrator privileges.

openclaw.ai supports a wide range of LLM providers:

• Cloud APIs: Anthropic, OpenAI, Google Gemini, OpenRouter, AWS Bedrock, and others
• Local LLMs: Ollama and LM Studio
• Self-hosted: custom OpenAI-compatible and native provider endpoints

You can even use multiple providers simultaneously and switch between them based on task requirements.

Yes! You can run openclaw.ai completely offline by using local LLMs such as:

• Ollama: automatic local model discovery on the default endpoint
• LM Studio: local OpenAI-compatible model serving
• Custom endpoints: manually configured self-hosted providers

This allows fully private operation without sending any data to external services.

The recommended updater detects the install type, runs diagnostics, and restarts the Gateway:

openclaw update

For Docker users:

export OPENCLAW_IMAGE="ghcr.io/openclaw/openclaw:latest"
./scripts/docker/setup.sh

This usually means Node.js isn't installed or the npm global bin isn't in your system PATH. Solutions:

1. Install a supported Node.js version (22.22.3+, 24.15+, or 25.9+) from nodejs.org
2. Check your version: node --version (must match 22.22.3+, 24.15+, or 25.9+)
3. If openclaw is not found, add npm global bin to PATH: export PATH="$(npm prefix -g)/bin:$PATH"

Try these solutions:

• Use the install script instead: curl -fsSL https://openclaw.ai/install.sh | bash
• Check your internet connection
• Temporarily disable antivirus/firewall
• If using VPN, try disconnecting

Don't use sudo npm install -g! Instead:

1. Fix npm permissions: mkdir ~/.npm-global && npm config set prefix '~/.npm-global'
2. Or use the install script: curl -fsSL https://openclaw.ai/install.sh | bash

To completely remove openclaw.ai:

# Uninstall the package
npm uninstall -g openclaw

# Remove configuration files
rm -rf ~/.openclaw

# For Docker
docker rm -f openclaw
docker rmi ghcr.io/openclaw/openclaw:latest