Deploy Your Assistant
Step-by-step guide to deploying your AI assistant in 60 seconds
One-Click Deployment
Deploying your personal AI assistant takes less than 60 seconds. Here's the complete flow:
Step 1: Choose a Plan
From your dashboard, click New Server. Select a plan that fits your needs:
| Plan | vCPU | RAM | Storage | Price |
|---|---|---|---|---|
| Starter | 2 | 4 GB | 40 GB | $49/mo |
| Professional | 4 | 8 GB | 80 GB | $99/mo |
| Enterprise | 8 | 16 GB | 160 GB | $199/mo |
Step 2: Add Your API Keys
Before deploying, make sure you've added your API keys in Settings:
- Anthropic API Key — for Claude-powered conversations
- OpenRouter API Key (optional) — for multi-model support
- Telegram Bot Token — connects your assistant to Telegram
Don't have a Telegram bot yet? See Telegram Setup.
Step 3: Hit Deploy
Click the Deploy button. MyClaw will:
- Provision a dedicated VPS in a European data center
- Install and configure OpenClaw
- Apply security hardening (firewall, fail2ban, non-root user)
- Connect your Telegram bot
- Run health checks to verify everything works
You'll see a real-time progress bar. Once complete, open Telegram and send your first message!
What Happens Behind the Scenes
Your server runs on Hetzner Cloud infrastructure in Germany or Finland. Each server is fully isolated — no shared resources with other users. We configure:
- UFW firewall with minimal open ports
- fail2ban for brute-force protection
- Non-root execution for the OpenClaw process
- Encrypted credential storage for your API keys
- Automatic updates managed by MyClaw
Troubleshooting
If deployment fails, click Retry Provision on the server page. Common issues:
- Invalid API key — double-check your keys in Settings
- Rate limiting — wait a moment and retry
- Temporary infrastructure issue — usually resolves within minutes