OpenClaw on DigitalOcean

Developer-friendly VPS with 1-Click Deploy — best for beginners

DigitalOcean is a popular choice for running OpenClaw—clear pricing, excellent documentation, and a 1-Click Deploy app in the marketplace that gets you running in minutes. New users get $200 in free credit, which covers several months of a basic Droplet. The UI is straightforward, support is responsive, and you're never far from a tutorial if something goes wrong.

This guide covers both the 1-Click Deploy path and manual setup. The 1-Click app is security-hardened (Docker isolation, unique gateway token, firewall rules, non-root execution, fail2ban). Manual setup gives you more control and is useful if you want to customize from scratch.

Option 1: 1-Click Deploy (Fastest)

  1. Create a DigitalOcean account (use the $200 credit promo if you're new)
  2. Go to Marketplace → search for OpenClaw
  3. Click Create OpenClaw Droplet
  4. Choose region—closest to you for SSH/WebChat; also consider where your LLM API runs (latency is often model round-trips, not the Droplet). Common picks: NYC, SFO, or Ashburn if you use US-based APIs.
  5. Select plan: 4GB RAM recommended (~$24/mo), 2GB (~$12/mo) works for light use
  6. Add SSH key for secure access
  7. Create Droplet

The 1-Click image comes pre-configured with OpenClaw, Docker, firewall, and a unique gateway token. After creation, SSH in and complete channel setup. Docs: DigitalOcean OpenClaw marketplace.

Option 2: Manual Setup

For full control, create a basic Droplet and install OpenClaw yourself:

  1. Create Droplet → Ubuntu 24.04 LTS
  2. Plan: Basic Droplet, $12/mo (2GB) or $24/mo (4GB)
  3. Add SSH key
  4. Create

Then SSH in and run:

Install OpenClaw
curl -fsSL https://openclaw.ai/install.sh | bash
npm install -g pm2
openclaw init
pm2 start 'openclaw gateway start' --name openclaw
pm2 save
pm2 startup

See VPS overview for firewall and reverse proxy setup.

Choosing a Droplet Size

  • 1 vCPU, 1GB RAM ($6/mo) — Absolute minimum. Fine for testing; may struggle with multiple channels or skills.
  • 1 vCPU, 2GB RAM ($12/mo) — Recommended for personal use. Comfortable for one user, few channels.
  • 2 vCPU, 4GB RAM ($24/mo) — Best balance. Handles teams, more channels, heavier skills.

Remember: LLM inference happens on Anthropic/OpenAI/Ollama servers. Your Droplet runs Node.js and the gateway—lightweight by comparison.

Why DigitalOcean for OpenClaw

  • Beginner-friendly — Simple UI, clear docs, predictable pricing
  • $200 free credit — New accounts; covers months of a basic Droplet
  • 1-Click Deploy — Pre-hardened OpenClaw image in the marketplace
  • Global regions — Deploy close to your users
  • Great support — Tutorials, community, responsive help

After install

  1. Run openclaw onboard for your LLM provider and first channel.
  2. Pick a walkthrough: channel hub or Telegram quickstart.
  3. Set pairing policy before you expose the bot to untrusted DMs.
  4. On a public VPS, keep the Gateway on loopback; use a reverse proxy only if you need remote WebChat (see remote access).

VPS pricing does not include LLM API spend—see monthly cost guide and example setups & model routing.

Related