How to Set Up OpenClaw on Hostinger VPS (Step-by-Step for Beginners)

I’ll be honest — when I first heard about OpenClaw, I rolled my eyes a little.

Another “AI assistant” tool. Another thing to configure. Another thing to break at 11pm when I just want to go to bed.

But here’s the thing. I spent a weekend actually setting this up on a Hostinger VPS, and it surprised me. Not because it’s magic — it’s not — but because the setup was genuinely simpler than I expected. No deep terminal work. No hunting Stack Overflow. No calling a developer friend.

If the phrase hostinger openclaw setup landed you here, you’re probably wondering the same thing I was: can a non-technical person actually get this running? The honest answer is yes. And I’m going to show you exactly how, step by step.

Before You Start — The Realistic Picture

Let me be straight with you before we get into it.

This isn’t a “set it up in 5 minutes” situation. More like 30 minutes if you move at a relaxed pace. It costs real money — the Hostinger KVM 2 plan (the minimum you need) runs roughly $5–$9/month depending on how long a billing cycle you commit to. And while you don’t need to know how to code, you do need to read carefully and follow each step in order.

If that sounds fair, let’s move.

What Is OpenClaw? (And Why Does It Need a VPS?)

OpenClaw used to be called Moltbot. Before that, Clawdbot. The name has changed a few times, but the idea has stayed the same: it’s a self-hosted AI assistant that lives on a server and talks to you through the messaging apps you already use — Telegram, WhatsApp, Slack, Discord, and a dozen others.

You give it an AI model to use (OpenAI’s GPT, Anthropic’s Claude, Google’s Gemini — your choice), and it becomes a personal assistant that can schedule tasks, automate workflows, answer questions, and execute actions — all through plain conversation.

Now, why a VPS? Simple. If you install OpenClaw on your laptop, it goes offline the moment you close the lid. A VPS is essentially a small computer that lives in a data center and never turns off. Your assistant stays available 24/7 even when you’re asleep or your internet goes down. Think of it as renting an always-on computer for about the cost of a Netflix subscription.

Hostinger offers a one-click OpenClaw template, which cuts what used to be a multi-hour manual install down to something a beginner can finish before lunch.

Step 1: Pick the Right Hostinger VPS Plan

The plan you choose matters more than people realize. OpenClaw runs inside Docker containers — essentially mini-programs inside your server — and if your server doesn’t have enough resources, things get sluggish fast.

For most people reading this, the KVM 2 plan is the right call. It has enough CPU and RAM to run OpenClaw comfortably, handle multiple messaging channels, and not break a sweat during normal use. If you’re planning to run several bots simultaneously or use OpenClaw’s browser automation features heavily, step up to KVM 4. But for getting started? KVM 2 is plenty.

One more thing while you’re on the plan page: you’ll see an option to add Nexos AI credits for $5. If you don’t already have API keys from OpenAI or Anthropic, grab this. It gives you immediate access to AI models without creating separate external accounts — saves a frustrating detour later.

On billing period — the 24-month plan has the lowest monthly rate. If you’re committed long-term, it’s the smart pick. If you want to test first, monthly works fine.

Step 2: Complete Your Purchase and Lock In the OpenClaw Template

Here’s where Hostinger’s setup earns its keep.

During checkout, you’ll reach a screen asking you to choose an application template. Find OpenClaw on VPS in that list — if you came from Hostinger’s dedicated OpenClaw page, it should already be selected. If not, just search for it in the catalog.

Also on this screen: the Nexos AI credits add-on and Oxylabs credits, which adds web search capability to your assistant. Both are optional. Neither is essential to get started.

Finish the purchase. You’ll land in your hPanel dashboard — that’s Hostinger’s control panel, where you’ll be spending the next 25 minutes or so.

Step 3: Configure Your VPS — Location, OS, and Backups

After purchasing, click Setup in your hPanel to start configuring the VPS. Three decisions here, and each one matters.

Server Location

Pick the region closest to where you physically are. Hostinger’s system usually auto-suggests the lowest-latency option — I’d trust that default. The closer the server to you, the snappier your assistant’s responses feel. I’m in Southeast Asia, so I stuck with the suggested server region rather than picking somewhere that sounded cooler.

Operating System

Go with Ubuntu 24.04 LTS. It has the longest support window of any current option, it’s the most stable, and it’s what OpenClaw’s Docker template is optimized for. Don’t overthink this one.

Daily Auto Backups

Turn this on. I’m going to say it louder: turn this on. OpenClaw can modify its own server environment — install new skills, reconfigure settings, change its own behavior. That’s powerful. It also means things can occasionally go sideways in ways that are hard to manually untangle. A daily snapshot means you’re never more than 24 hours away from a clean restore point.

Step 4: Install Docker Manager on Your VPS

Docker is what actually runs OpenClaw behind the scenes. Without getting too technical — Docker creates isolated environments called containers. Each container runs independently without interfering with anything else on the server. It’s the reason the one-click install is even possible.

In your hPanel, find Docker Manager in the left sidebar.

1. If you ticked the Docker Manager option during setup, it’s already installed. Click in and confirm it shows a Running status.
2. If it’s not installed yet, click Install and give it 2–3 minutes.
3. Once Docker Manager is running, click into the Catalog tab — that’s where you’ll find OpenClaw in the next step.

Step 5: Deploy OpenClaw from the Docker Catalog

This is the step I was most nervous about. It turned out to be the easiest one.

Inside Docker Manager, go to Catalog and search for OpenClaw. Click Install.

A form will pop up asking for environment variables. That sounds complicated — it’s basically just a settings page. Here’s what to fill in:

1. OPENCLAW_GATEWAY_TOKEN — Auto-filled by the system. Do not edit it. Do, however, copy it right now and paste it somewhere safe. Your notes app, your password manager — anywhere. Don’t close this screen without saving that token. It’s the only key to your dashboard.
2. API Keys (OpenAI / Anthropic / Gemini / xAI) — If you have your own API keys, paste them in the corresponding fields. If you went with Nexos AI credits, leave these blank.
3. WHATSAPP_NUMBER — Enter your number if WhatsApp is your preferred channel. You can also skip this and configure it later.

Once your gateway token is saved, click Deploy. The deployment takes 2–5 minutes. When the status indicator flips to a green Running status, you’re good to go.

Step 6: Connect an AI Model With Your API Key

OpenClaw is the shell. The AI model is the brain inside it. Right now your assistant is running but has nothing to think with. You need to connect it to an AI provider using an API key.

If you bought Nexos AI credits: Skip this step entirely. Your AI access is already linked through your Hostinger account.

If you’re using your own API key (OpenAI example):

1. Go to platform.openai.com and log in.
2. Navigate to API Keys in the left sidebar.
3. Click Create new secret key. Give it a name — I called mine “OpenClaw VPS.”
4. Copy the key immediately. You will not see it again after closing the screen.
5. Back in Hostinger’s Docker Manager, open your OpenClaw project settings and paste the key into the OPENAI_API_KEY field.
6. Restart the OpenClaw container to apply the change.

Getting an Anthropic or Gemini key works the same way — each provider has their own dashboard, but the process is nearly identical.

⚠️ Important: Treat API keys exactly like passwords. Don’t paste them in chat messages. Don’t post them in public forums. A leaked key can rack up charges on your account fast.

Step 7: Access the OpenClaw Dashboard With Your Gateway Token

If you saved your gateway token back in Step 5, this part is fast.

In Docker Manager, find your OpenClaw project and click the Open link — usually a purple link in the lower left of the project tile. This takes you to your OpenClaw dashboard, which runs at your VPS’s IP address on port 18789.

Paste your saved token into the login field and click Authenticate.

If you lost your token, open the browser-based terminal in hPanel and run:

cat ~/.clawdbot/.env | grep CLAWDBOT_GATEWAY_TOKEN

The token will appear in the output. The “clawdbot” naming in the files is just the old name from before the rebrand — everything works fine, don’t let it throw you off.

Once inside the dashboard, confirm the system shows a Healthy status. OpenClaw will then run you through a short onboarding interview — your assistant’s name, communication preferences, a bit of personalization. Takes about 2 minutes. Do it properly; it shapes how your assistant behaves day to day.

Step 8: Connect Telegram or WhatsApp

You’ve got a running AI assistant. Now give it a way to hear from you.

Connecting Telegram

1. Open Telegram and search for @BotFather.
2. Send the message /newbot and follow the prompts to name your bot and create a username.
3. BotFather gives you an API token at the end — copy it.
4. In your OpenClaw dashboard, go to Channels → Telegram.
5. Paste the token and click Connect.
6. Open your new bot in Telegram and send “Hi.” OpenClaw will respond with a short setup conversation.

Connecting WhatsApp

1. In your OpenClaw dashboard, go to Channels → WhatsApp.
2. A QR code appears on screen.
3. On your phone, open WhatsApp → Settings → Linked Devices → Link a Device and scan the QR code.
4. Send a test message — your assistant should respond within seconds.

That’s it. OpenClaw is live on your Hostinger VPS, online 24/7, responding to your messages whether you’re at your desk or somewhere without your laptop.

Three Mistakes I See Beginners Make

1. Skipping the daily backup toggle during setup. People say “I’ll enable that later.” They don’t. Then OpenClaw reconfigures something, something breaks, and they’re stuck with no restore point. Thirty seconds during Step 3 saves hours of pain later.

2. Not saving the gateway token before closing the deployment screen. The most common one by far. The token appears once during deployment and then you have to dig it out through the terminal. Avoidable. Save it before you click anything else in Step 5.

3. Overriding the server location suggestion. Picking a server in London when you’re based in Dhaka means noticeably slower response times on every single message. Latency is real. Pick the closest region, not the most impressive-sounding one.

Frequently Asked Questions

What is OpenClaw and how does it actually work?

OpenClaw is a self-hosted AI assistant platform — formerly known as Moltbot and Clawdbot — that runs on your own server. It connects to AI models like OpenAI’s GPT or Anthropic’s Claude and lets you interact with that AI through messaging apps you already use, like Telegram and WhatsApp. Because it runs on a VPS, it stays active around the clock regardless of whether your personal devices are on.

Do I need any coding experience to set up OpenClaw on Hostinger?

No. Hostinger’s one-click Docker template handles the server-side setup automatically. You fill in a few fields in a web form, click Deploy, and the platform handles the technical work. This guide was written specifically for people with no server or coding background.

What is the cheapest Hostinger plan that can run OpenClaw properly?

The KVM 2 plan is the minimum recommended configuration. It provides enough CPU and memory to run OpenClaw’s containers reliably for personal use, including multiple connected messaging channels.

How do I recover my OpenClaw gateway token if I lose it?

Open the browser-based terminal in hPanel and run: cat ~/.clawdbot/.env | grep CLAWDBOT_GATEWAY_TOKEN — the token will appear in the output. The config files still use the legacy “clawdbot” naming during the ongoing rebrand, but everything functions normally.

Can OpenClaw connect to Telegram and WhatsApp at the same time?

Yes. OpenClaw supports simultaneous connections to multiple channels — including Telegram, WhatsApp, Slack, Discord, Google Chat, Signal, iMessage, and Microsoft Teams. All active channels are managed from the Channels section in your dashboard.

Want the Setup Checklist PDF?

I put together a one-page PDF that condenses every step in this guide into a quick reference — including the exact fields to fill in during deployment and the terminal command for recovering a lost gateway token.

Drop your email below and I’ll send it straight to your inbox. No sequences, no sales emails. Just the checklist.

I need this.

Wrapping Up — What You Should Do Right Now

You’ve covered everything: choosing the right plan, deploying the template, connecting an AI model, and linking your first messaging channel. OpenClaw on Hostinger is genuinely one of the more accessible self-hosted AI setups I’ve tried — and I’ve tried a few.

The one thing that catches people out every time isn’t a technical problem. It’s hesitation. The setup looks intimidating from the outside and turns out to be straightforward once you actually start.

So here’s your first action: click the link below, land on Hostinger’s OpenClaw VPS page, and grab the KVM 2 plan. You can have your assistant live and responding to your first message in under 30 minutes.