The Ultimate Self-Hosted AI Agent: Setting Up OpenClaw on a Mac Mini
Deploying OpenClaw on a Mac Mini creates a high-performance, autonomous AI workstation. To breathe life into this agent, you need to connect it to an LLM "brain" via an API.
1. Issuing Your LLM API Keys
OpenClaw requires a connection to a model provider. Follow these steps to get your keys:
- Anthropic (Recommended): Go to the Anthropic Console, create an account, and generate a
sk-ant-...key. Claude 3.5 Sonnet is highly recommended for its balance of speed and reasoning. - OpenAI: Visit the OpenAI Platform, navigate to the API Keys section, and issue a
sk-...key.
2. Installation & Daemon Setup
Open your terminal and execute the following to install the service:
# Install OpenClaw globally
npm install -g openclaw@latest
# Run the onboarding wizard
openclaw onboard --install-daemon
3. Registering API Keys in OpenClaw
During the onboard process, the wizard will ask for your provider and key. If you need to add or change them later, follow this guide:
Configuration Path:
Open the hidden configuration file in your home directory:
Open the hidden configuration file in your home directory:
~/.openclaw/.env
# Example .env configuration
ANTHROPIC_API_KEY=your_anthropic_key_here
OPENAI_API_KEY=your_openai_key_here
4. Features & Operating Costs
Vibe Coding
The agent executes terminal commands and organizes system files based on your natural language instructions.
Budgeting
While hosting is nearly free (approx. $1/mo electricity), expect API token costs ranging from $10 to $40+ per month depending on task frequency.
Conclusion
Setting up your API keys is the final step in activating your autonomous worker. With OpenClaw and a Mac Mini, you now have a private, secure AI server ready to handle your daily digital drudgery.

댓글
댓글 쓰기