OpenClaw Setup Guide | aiwithblake.com

1

Install OpenClaw

Open your terminal and run the installation command for your platform:

macOS (Homebrew):

brew install openclaw

Linux:

curl -fsSL https://openclaw.ai/install.sh | sh

Windows (PowerShell):

irm https://openclaw.ai/install.ps1 | iex

Verify installation by running: openclaw --version

2

The Gateway is the core service that manages your AI connections:

openclaw gateway start

This starts the OpenClaw daemon. Keep this terminal window open, or run with --background flag.

3

Pull a lightweight model to get started:

openclaw model pull llama3.2

This downloads the Llama 3.2 model (great for general use). For coding, try codellama. For larger context, try llama3.2:70b (requires more RAM).

4

Start a conversation with your local AI:

openclaw chat

Type a message and press Enter. Your AI is now running completely locally.

Press Ctrl+C to exit the chat interface.

5

To use your AI through Discord:

# Configure Discord bot
openclaw channel add discord

# Follow prompts to enter your bot token
# Your AI is now available in Discord

You'll need a Discord bot token. Create one at discord.com/developers

6

To use your AI through Telegram:

# Configure Telegram bot
openclaw channel add telegram

# Follow prompts to enter your bot token
# Chat with your AI on Telegram

Get a bot token from @BotFather on Telegram

7

You're all set! Here are some things to try:

!

Gateway won't start? Check if port 8080 is already in use: lsof -i :8080

Model download fails? Ensure you have enough disk space (models are 2-5GB each).

Slow responses? Larger models need more RAM. Try a smaller model or enable GPU acceleration.

Need help? Email me at blake@aiwithblake.com