Setting up OpenClaw on Ubuntu takes about twenty minutes if you follow the right order of steps and a lot less if your server already has everything preloaded.

The challenge is that many installation guides leave out important details. One assumes you already have the correct version of Node.js, another skips system preparation, and before long you’re troubleshooting errors instead of getting your AI assistant online.

So we made sure to go through every necessary step in the right order with detailed explanations to solve that.

This guide walks you through the complete deployment process, from connecting to your Ubuntu VPS to accessing the OpenClaw dashboard securely.

Along the way, you’ll also see where a pre-configured OpenClaw VPS can save time by removing several manual setup steps.

Now, before getting started, make sure you have the following:

• An Ubuntu VPS running Ubuntu 20.04, 22.04, or 24.04 with root SSH access
• An AI API key from OpenAI, OpenRouter, or another supported provider
• A Telegram bot token if you plan to message OpenClaw through Telegram

With these ready, let’s begin the deployment.

Steps to Deploy OpenClaw on an Ubuntu VPS

Step 1: Choose the Right Ubuntu VPS for OpenClaw

Your VPS is the foundation of your OpenClaw deployment, so it’s worth choosing one that meets the recommended requirements from the start.

At a minimum, you’ll need an Ubuntu VPS running Ubuntu 20.04, 22.04, or 24.04 with root SSH access so you can install software and configure system services.

On the hardware side, OpenClaw’s gateway and background daemon need breathing room. A 1 vCPU, 2 GB RAM minimum is the practical floor, and anything below that and you risk the Node.js install timing out or the daemon restarting itself mid-conversation.

If you are deploying through Truehost, our OpenClaw VPS plans start at that spec, with full root access from first login.

Better still, if you choose our pre-configured OpenClaw instance, Steps 2 and 3 below are already done for you. The server ships with Ubuntu, Node.js, and OpenClaw already installed, so you go straight to onboarding.

Step 2: Connect and Prepare Your Ubuntu VPS

Once your Ubuntu VPS is ready, log into your server as root:

ssh root@your-server-ip

After logging in, update your system packages before installing anything else. Running OpenClaw on top of outdated system libraries is an easy way to introduce silent dependency conflicts:

apt update && apt upgrade -y

Then install the prerequisites OpenClaw and Node.js will need during setup:

apt install -y curl git build-essential

Here’s what each package is used for:

• curl downloads the official OpenClaw installer.
• git supports repositories and development tools that OpenClaw may use.
• build-essential provides the compiler and build tools required by some Node.js packages.

Once these commands finish, your Ubuntu VPS is ready for the OpenClaw installation.

If you’re using a pre-configured Truehost OpenClaw VPS, this step has already been completed for you. The server is provisioned with the required system packages, so you can move directly to installing OpenClaw.

Step 3: Install OpenClaw on Your Ubuntu VPS

The easiest way to install OpenClaw is by using the official installer script.

It automatically detects your operating system, installs a compatible version of Node.js if needed, installs OpenClaw, and then launches the onboarding wizard, all from a single command:

curl -fsSL https://openclaw.ai/install.sh | bash

If you prefer to install first and run onboarding separately, add the --no-onboard flag:

curl -fsSL https://openclaw.ai/install.sh | bash -s -- --no-onboard

After the installation completes, verify that everything is working correctly:

openclaw --version
openclaw doctor

The first command confirms that OpenClaw was installed successfully, while the second checks your environment for any missing dependencies or configuration issues.

If you already manage your own Node.js installation and prefer not to use the installer script, you can install OpenClaw manually with npm:

sudo npm install -g openclaw@latest

Either path ends at the same place: OpenClaw installed on your Ubuntu VPS and ready for onboarding.

Step 4: Run the OpenClaw Onboarding Wizard

If you ran the installer script without the --no-onboard option, the onboarding wizard starts automatically after the installation finishes. You can simply follow the prompts and continue with the setup.

If you used the --no-onboard flag or installed OpenClaw manually with npm, start the onboarding wizard yourself.

openclaw onboard --install-daemon

This command does more than configure OpenClaw. It also installs the OpenClaw gateway as a systemd user service, allowing it to continue running in the background after the setup is complete.

During onboarding, you’ll be guided through a series of prompts to configure your deployment:

1. Review the risk acknowledgment and confirm that you want to continue.
2. Select the Quick Start option for the fastest setup.
3. Enter your AI API key from OpenAI, Google AI Studio, OpenRouter, or another supported provider.
4. Choose a messaging platform if you want to control OpenClaw remotely.

If you select Telegram, the wizard will ask for your bot token. You can create one by chatting with BotFather on Telegram and following the prompts to generate a new bot.

Once you receive the token, paste it into the onboarding wizard to complete the connection.

After those four steps are complete, OpenClaw is configured and linked to both your AI provider and your messaging channel.

Step 5: Run OpenClaw as a Background Service

An AI agent that stops when your SSH session closes is not particularly useful. To keep OpenClaw running after you log out, enable lingering for your user account. This tells Ubuntu to keep your background services alive without an active session:

loginctl enable-linger $USER

Next, confirm that the OpenClaw gateway is running:

openclaw gateway status

You should see that the gateway service is active. If you ever need to troubleshoot your deployment or monitor activity in real time, view the service logs with the following:

journalctl --user -u openclaw -f

These logs are especially useful for diagnosing connection issues, checking gateway activity, or confirming that OpenClaw has started successfully after a reboot.

This is also where hosting quality becomes directly relevant to how your agent actually performs.

Since OpenClaw is designed to run continuously, a stable hosting environment helps keep your gateway online and responsive.

Features such as reliable infrastructure, isolated resources, and regular backups can help reduce unexpected downtime and keep your AI assistant available around the clock.

Our Truehost OpenClaw plans include isolated vCPU allocation, daily SSD snapshots, and built-in firewall protection by default.

Step 6: Access the OpenClaw Dashboard Securely

Once OpenClaw is running, you can access its web dashboard to monitor your deployment, check the gateway status, and manage your AI assistant.

Start the dashboard on your Ubuntu VPS by running:

openclaw dashboard

For security, the dashboard listens only on localhost by default. This means it isn’t exposed directly to the internet, helping protect your deployment from unauthorized access.

To open the dashboard on your own computer, create an SSH tunnel from your local machine:

ssh -N -L 18789:localhost:18789 username@your-server-ip

Leave that terminal running, then open your browser and go to:

http://localhost:18789

Your dashboard loads locally, while the actual process runs on your VPS. The tunnel keeps the connection encrypted without exposing any port to the public internet.

From the dashboard, you can verify that your AI provider, gateway, and messaging platform are connected correctly.

OpenClaw on Ubuntu VPS FAQs

The Faster Way: Deploying OpenClaw on a Pre-Configured Ubuntu VPS

Everything in this guide works, and knowing the manual process is worth your time if you want full visibility into your own stack.

That said, if the goal is simply to have OpenClaw running today, Truehost’s OpenClaw VPS hosting removes the setup overhead entirely. Our plans deploy with Ubuntu, Node.js, and the OpenClaw orchestration engine already installed and configured, and Playwright is included.

You still get full root SSH access from the first login. Nothing about control or flexibility changes, just that you skip the parts that usually eat up an afternoon.

Plan options:

• OpenClaw Starter: 1 vCPU, 2 GB RAM, 50 GB NVMe storage. Suitable for a single agent with standard workloads.
• OpenClaw Pro: 2 vCPUs, 4 GB RAM, 100 GB NVMe storage. Better for heavier browser automation.
• OpenClaw Business: 4 vCPUs, 8 GB RAM, 200 GB NVMe storage. Great for multiple connected channels.

All our plans include root access, daily backups, isolated resources, and firewall protection. Storage, RAM, and vCPU can be scaled later from your dashboard without downtime, so you are not locked into your initial spec as your agent’s workload grows.

Now, if you’re ready to start your deployment journey, get your Truehost Openclaw VPS hosting now and begin right away.