Hetzner vs Linode

• A server that stays online OpenClaw works best on a VPS or machine that can run continuously without stopping.
• Enough RAM A very small server can struggle during setup and regular use. More RAM gives OpenClaw enough room to run properly.
• Enough disk space You need storage for the install, system files, logs, and other working data.
• An AI model provider OpenClaw needs a provider like OpenAI or Anthropic to generate replies and complete tasks.
• A secure setup If you host it on a VPS or public server, you need to protect access and avoid leaving it exposed.

• Dashboard experience
  Linode has a more polished dashboard, so it usually feels easier to use at first. Hetzner is more simple and direct, but it feels more basic.
• Server setup
  Both platforms make it fairly easy to launch a VPS. The bigger difference starts after the server is live.
• Backups and networking
  Linode feels easier if you want more cloud tools in one place. Hetzner gives you the main features you need, but in a more stripped-down way.
• Storage and extra services
  Linode makes more sense if you may want more services around your VPS later. Hetzner is better if you mainly want a VPS and do not want to pay extra for more platform features.
• Day-to-day management
  Linode feels more comfortable if you expect to manage several things over time. Hetzner works well if you mainly want simple server management at a lower cost.
• For OpenClaw
  Linode is easier if you want a more comfortable cloud platform. Hetzner is better if you want lower-cost self-hosting. Neither platform removes the server work, so you still need to handle setup, updates, security, and maintenance yourself.

Buying a VPS does not mean OpenClaw is ready. You still have to set it up and keep it running.

• Log in and secure the server
  Add SSH keys, lock down access, and keep the server private where possible.
• Install OpenClaw
  Run the installer: curl -fsSL https://openclaw.ai/install.sh | bash on Linux/macOS, or use the PowerShell installer on Windows.
• Run onboarding
  You still need to configure the setup with openclaw onboard --install-daemon. This handles model auth, gateway setup, channels, daemon install, and health check.
• Add your AI provider API key
  OpenClaw still needs a provider like OpenAI or Anthropic connected before it can work.
• Set up the gateway
  OpenClaw needs the gateway installed and running in the background for ongoing use. Official docs point to openclaw onboard --install-daemon or openclaw gateway install, and you can verify with openclaw gateway status.
• Connect your channels
  If you want Telegram, WhatsApp, Discord, or other chat channels, you still need to configure them after install.
• Open the dashboard and test it
  You still need to open the Control UI, check that the gateway is running, and send a test message.
• Keep it updated
  OpenClaw has a separate update flow, so updates and maintenance are still part of self-hosting.

Managed hosting means you do not manage the VPS, Docker, updates, or uptime yourself.

• Deploy the agent
  Start the agent from the dashboard instead of setting up a VPS manually.
• Add your AI model API key or use included credits
  The agent still needs model access to work, whether through your own API key or a hosting plan that includes credits.
• Connect your communication channels
  Link the channels where you want to use the agent, such as Telegram, WhatsApp, or Discord.
• Set the agent’s role and tasks
  Decide what the agent should do, how it should respond, and what kind of work it should handle.
• Enable the skills you want
  If you need extra capabilities, you still choose and enable those skills as part of setup.
• Test the agent
  Send a few messages and make sure everything works the way you expect.

What You Skip With Managed Hosting

• No VPS setup
• No Docker setup
• No firewall setup
• No server updates
• No uptime maintenance