The easiest way to run OpenClaw.
No Mac Mini. No Docker config. No port forwarding. One command, and you're in.
$ npm install -g @machine0/cli
$ machine0 use openclaw
✓ VM is starting (large, openclaw image)
✓ Connected
[ ubuntu@openclaw ] ~ # Why cloud?
OpenClaw needs a machine that's always on, always connected, and has enough power to run an LLM gateway plus your agent workflows. Most guides tell you to buy a Mac Mini M4 for $500-700.
A cloud VM is better. No upfront cost. Spin up in seconds. Destroy when you're done. Run it in any region. And you don't have a $700 computer sitting under your desk that you need to keep updated.
Mac Mini vs machine0
| Mac Mini M4 | machine0 | |
|---|---|---|
| Cost | $500-700 upfront | $0.052/hr (~$38/mo) |
| Setup | Hours (OS, Docker, config, networking) | 30 seconds |
| Management | You maintain it | We handle it |
| Location | Your desk | US, UK, EU, or Asia |
| Scale | Fixed hardware | Resize anytime |
| Cleanup | Sell it on eBay | machine0 rm |
What is OpenClaw?
The fastest-growing open-source project on GitHub. 247,000+ stars in 60 days.
OpenClaw is an autonomous AI agent you talk to through WhatsApp, Telegram, Discord, or Signal. It connects to an LLM - Claude, GPT, or DeepSeek - and executes tasks on your machine: writing code, managing files, browsing the web, running shell commands, and orchestrating complex multi-step workflows.
13,700+ community-built skills. Persistent memory across sessions. Runs as a background daemon with a heartbeat cron.
Pricing
We recommend large (2 vCPU, 4 GB RAM, $0.052/hr) for OpenClaw. Works with medium too if you're just getting started.
| Size | CPU | RAM | Disk | $/hour | ~$/month |
|---|---|---|---|---|---|
| medium | 2 vCPU | 2 GB | 60 GB | $0.026 | $19 |
| large | 2 vCPU | 4 GB | 80 GB | $0.052 | $38 |
| xl | 4 vCPU | 8 GB | 160 GB | $0.104 | $76 |