How to Self-Host OpenClaw AI Agent on a VPS: Complete 2026 Guide

OpenClaw (formerly Clawdbot/Moltbot) is an open-source AI assistant that actually does things — not just chat. It runs 24/7 on your own VPS, connects to WhatsApp, Telegram, Discord, Slack, and 50+ other platforms, and lets you choose your AI model: Claude, GPT, Gemini, or even local models like Llama and Mistral.

Unlike cloud AI subscriptions, you self-host OpenClaw on your own server. You pay a flat monthly VPS fee and get unlimited conversations with full data privacy.

In this guide, we'll cover everything: what OpenClaw can do, how to choose a hosting provider, and how to get it running.

---

What Can OpenClaw Do?

OpenClaw isn't just a chatbot — it's a full AI agent platform:

• Web browsing — research topics, extract data, monitor websites
• File management — create, edit, organize files on your server
• Command execution — run scripts, manage services, automate DevOps tasks
• Task scheduling — set up recurring automations, reminders, and monitoring
• Messaging integration — WhatsApp, Telegram, Discord, Slack, Signal, iMessage, Microsoft Teams
• 50+ platform integrations — GitHub, Spotify, cloud storage, smart devices, CI/CD pipelines
• Multiple AI models — Claude, GPT, Gemini, Llama, Mistral, and more

All of this runs on your server, meaning your data never leaves your infrastructure unless you explicitly connect it to external services.

---

Minimum VPS Requirements

Component	Minimum	Recommended
CPU	2 vCPU	4+ vCPU
RAM	4 GB	8+ GB
Storage	40 GB SSD	75+ GB NVMe
OS	Ubuntu 22.04+ / Debian 12+	Ubuntu 24.04
Network	1 TB bandwidth	Unlimited preferred

For personal use, 4 vCPU + 8 GB RAM is the sweet spot. If you plan to run local AI models alongside OpenClaw, go for 16 GB+ RAM.

---

Best VPS Providers for OpenClaw (2026)

We've compared six providers that support OpenClaw hosting. Here's how they stack up:

1. Hostinger — Best for Beginners (1-Click Install)

Hostinger offers 1-click OpenClaw installation through their control panel, making it the easiest way to get started.

Plan	vCPU	RAM	Storage	Price
KVM 1	1	4 GB	50 GB NVMe	$6.49/mo
KVM 2	2	8 GB	100 GB NVMe	$8.49/mo
KVM 4	4	16 GB	200 GB NVMe	$14.49/mo

Pros: 1-click install, AI Assistant in panel, 24/7 support, global data centers Cons: No hourly billing, weekly (not daily) backups

2. Contabo — Most Resources per Dollar

Contabo offers the most raw resources at the lowest price point — ideal if you want to run local AI models alongside OpenClaw.

Plan	vCPU	RAM	Storage	Price
Cloud VPS 10	4	8 GB	75 GB NVMe	€4.50/mo
Cloud VPS 20	6	12 GB	100 GB NVMe	€7.00/mo
Cloud VPS 40	12	48 GB	250 GB NVMe	€25.00/mo

Pros: Unbeatable resources/price, unlimited traffic, 9 data center regions, DDoS protection Cons: No 1-click install, monthly minimum contract

3. SSD Nodes — Best Value for Large Plans

SSD Nodes provides dedicated Intel Silver CPUs with generous resources, especially on larger plans.

Plan	vCPU	RAM	Storage	Price
KVM/SMALL	2	8 GB	160 GB SSD	$5.50/mo
KVM/MEDIUM	4	16 GB	320 GB SSD	$9.42/mo
KVM/2X-LARGE	8	32 GB	480 GB SSD	$11.08/mo

Pros: Dedicated Intel Silver CPUs, 14-day money-back guarantee, 99.9% uptime, 12 global locations Cons: No 1-click install, 3-year billing for best prices

4. Amazon Lightsail — Enterprise & AWS Ecosystem

AWS introduced OpenClaw on Lightsail for users who want to run AI agents within the AWS ecosystem. Best for teams already using AWS services.

Pros: AWS infrastructure reliability, easy integration with Bedrock and other AWS services, predictable pricing Cons: Higher cost than budget VPS providers, AWS learning curve

5. xCloud — Managed OpenClaw Hosting

xCloud offers a fully managed hosting experience where they handle the server setup and maintenance for you.

Pros: No technical knowledge required, managed updates, 5-minute deployment Cons: Less control than raw VPS, higher cost for managed service

6. TrueNAS — Self-Host at Home

If you have a TrueNAS server at home or in your office, you can deploy OpenClaw directly from the TrueNAS Apps catalog.

Requirements: 2 CPU cores, 4 GB RAM minimum Supports: Token, password, or trusted proxy authentication Storage: Flexible — ixVolume, host paths, SMB/CIFS, NFS mounts

Pros: No monthly hosting cost, full local control, community-maintained Cons: Requires existing TrueNAS hardware, you manage everything

---

Quick Comparison Table

Provider	Starting Price	vCPU	RAM	1-Click Install	Best For
Hostinger	$6.49/mo	1	4 GB	Yes	Beginners
Contabo	€4.50/mo	4	8 GB	No	Max resources
SSD Nodes	$5.50/mo	2	8 GB	No	Large plans
AWS Lightsail	~$10/mo	2	2 GB	Yes	AWS users
xCloud	Varies	—	—	Managed	Non-technical users
TrueNAS	Free	2	4 GB	App catalog	Home servers

---

How to Set Up OpenClaw

Option A: 1-Click Install (Hostinger)

1. Sign up for a Hostinger VPS plan
2. In the control panel, find 1-click OpenClaw under AI & Automation
3. Click Install — the system handles Docker, networking, and configuration
4. Connect via SSH to run the onboarding wizard
5. Add your AI API keys (Claude, GPT, or Gemini)
6. Connect your messaging platforms (WhatsApp, Telegram, etc.)

Option B: Manual Install (Any VPS)

For providers like Contabo, SSD Nodes, or any Linux VPS:

# Update system
sudo apt update && sudo apt upgrade -y

# Install Docker
curl -fsSL https://get.docker.com | sh
sudo usermod -aG docker $USER

# Deploy OpenClaw
docker run -d \
  --name openclaw \
  --restart unless-stopped \
  -p 3000:3000 \
  -v openclaw-data:/app/data \
  openclaw/openclaw:latest

Then open http://your-server-ip:3000 to access the setup wizard.

Option C: TrueNAS App

1. Open TrueNAS Scale dashboard
2. Go to Apps → Available Applications
3. Search for "OpenClaw"
4. Click Install, configure resources (2 CPU, 4 GB RAM minimum)
5. Access via the assigned port (default: 30262)

---

Choosing Your AI Model

OpenClaw supports multiple AI providers. Here's how to decide:

Model	Best For	API Cost
Claude (Anthropic)	Complex reasoning, long documents, coding	Per token
GPT (OpenAI)	General tasks, creative content	Per token
Gemini (Google)	Multimodal tasks, Google ecosystem	Per token / free tier
Llama (Meta)	Privacy-first, no API cost	Free (local)
Mistral	European data compliance, efficiency	Per token / local

---

Security Best Practices

Since OpenClaw has access to execute commands and manage files, security is critical:

1. Always use SSH tunnels for the Control UI — don't expose port 3000 publicly
2. Enable token authentication (recommended over password auth)
3. Keep Docker and OpenClaw updated regularly
4. Use a firewall — only open ports you actually need
5. Run on a dedicated VPS — SSD Nodes recommends isolating OpenClaw from your main machine for security

---

Cost Comparison: Cloud AI vs Self-Hosted OpenClaw

	Cloud AI (ChatGPT Plus)	Self-Hosted OpenClaw
Monthly cost	$20/mo per user	$4.50–$6.49/mo total
Conversations	Limited	Unlimited
Custom integrations	Limited	50+ platforms
Data privacy	Cloud-stored	Your server only
Always-on agent	No	Yes, 24/7
Multiple users	Extra cost	Included

Self-hosting OpenClaw on a €4.50/mo Contabo VPS gives you an always-on AI agent for less than a quarter of the price of ChatGPT Plus — with unlimited conversations and full privacy.

---

Getting Started

1. Pick a provider — Hostinger for easiest setup, Contabo for best resources
2. Deploy OpenClaw — 1-click or Docker install
3. Connect your AI — add API keys for Claude, GPT, or Gemini
4. Connect your messaging — WhatsApp, Telegram, Discord, or Slack
5. Start automating — let your AI agent handle tasks 24/7

Your personal AI agent is one VPS away. Pick a provider, deploy OpenClaw, and let it work for you around the clock.