Sign In →
📖 Setup Guide

Up and running in
under an hour

Three steps to running AI agents 24/7 on your own hardware — no cloud, no monthly fees, no babysitting.

1 Buy a Mac mini
2 Install your AI runtime
3 Install your agent
01

Buy a Mac mini

The Mac mini is the perfect always-on agent host. Silent, efficient, and powerful enough to run dozens of bots simultaneously.

Recommended
Mac mini M4 Pro
Starting at $1,399

The M4 Pro has 14 CPU cores and handles parallel workloads (multiple bots, AI inference, a live dashboard) without breaking a sweat. 24GB unified memory is the sweet spot.

Buy on Apple.com →
Budget pick
Mac mini M4
$599

Fine for 1–3 bots. 16GB RAM. Will feel tight if you add more workloads later.

Power pick
Mac mini M4 Pro (24GB)
$1,399

What we use. Handles 10+ bots, a full dashboard, and AI model calls without hesitation.

Overkill
Mac Studio M4 Max
$1,999+

For running local LLMs alongside your bots. Unnecessary unless you want to self-host AI models.

💡
Why not a server or cloud VM?
Mac mini uses ~6W idle — less than a lightbulb. No monthly fees, no latency overhead, and macOS runs your bots natively with zero configuration drama. One-time hardware cost beats cloud bills within months.

Setup checklist

02

Install your AI runtime

Your agents are powered by Claude (Anthropic's AI). OpenClaw is the runtime that connects Claude to your Mac — handling scheduling, memory, notifications, and tool access.

Install Homebrew

Homebrew is the package manager for macOS. Open Terminal and run:

bash
/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)"

Install Node.js and OpenClaw

bash
brew install node
npm install -g openclaw

Get your Claude API key

1 Go to console.anthropic.com
2 Create an account and add a payment method
3 Go to API Keys → create a new key
4 Copy it — you'll need it in the next step

Initialize OpenClaw

bash
openclaw init

This walks you through connecting your Claude API key, setting up your agent workspace, and optionally linking Telegram or iMessage for notifications.

Start the gateway

bash
openclaw gateway start
✅
Verify it's working
Run openclaw status — you should see the gateway running and your agent ready. You can now chat with Claude directly from Terminal or Telegram.
03

Install your agent

Each agent on 49ents comes as a ready-to-run Python package with a setup script. Most are running within 5 minutes of download.

Prerequisites

bash
brew install [email protected]
pip3 install -r requirements.txt

General install flow

bash
# Example for a trading bot
cd ~/agents/kraken-momentum/
cp .env.example .env
nano .env          # paste your Kraken API key + secret
python3 setup.py --install
# → Installs as a macOS LaunchAgent
# → Starts automatically on login and after reboot

Verify it's running

bash
tail -f ~/agents/kraken-momentum/bot.log
📡
Remote access
SSH into your Mac mini from anywhere: ssh yourusername@your-mac-ip. Pair with Cloudflare Tunnel to expose your dashboard publicly with a real domain — no port forwarding needed.
🤖
Let your AI manage everything
With OpenClaw running, your Claude agent can monitor bot logs, restart crashed services, and alert you via Telegram — all without you touching the Mac mini. Message it: "Check on my trading bots" and it'll report back in seconds.

Troubleshooting

Bot isn't starting after reboot
Check launchctl list | grep your-agent-name. If missing, re-run python3 setup.py --install.
API authentication errors
Double-check your .env file — no extra spaces or quotes around keys. Ensure your API key has the correct permissions on the exchange/service.
Need help?
Join our Discord community or email us at [email protected]. Most issues resolved within an hour.

Ready to deploy?

Browse our agent library and get your first bot running today.

Browse Agents →