A couple of weeks ago OpenClaw went completely viral. Everyone started discovering this open source tool and finding a ton of super creative uses for it. A free tool that more than one company would have loved to sell and offer to their clients.
Now that the hype has passed, we can look at better use cases more calmly. In this post I’m going to show you how to install OpenClaw the easiest and quickest way, without any complications. And heads up, this is aimed at less technical people, all those folks who are having trouble configuring it, find steps unclear, and don’t really understand what they’re doing.
So if you’re one of them or you simply want to install it quickly, keep reading. Everything is here, well explained, no redirecting you anywhere else.
What is OpenClaw? (Quick recap)
For those who still don’t know, OpenClaw (formerly MoltBot, and before that ClawdBot) is an open source assistant that you can connect to any LLM, meaning any artificial intelligence.
Simply with natural language, through a chat, you can set up scheduled tasks. For example, what people do with N8N, but instead of building visual flows, you just type it out and ask.
And not only that, it’s capable of executing and performing any action you teach it. Teaching it is also simple: you just write and ask. Many people are defining it as a personal AI secretary.
Note: If you want a more detailed explanation of what OpenClaw is, I have a dedicated post about it.
Where is OpenClaw installed?
OpenClaw can be installed on any machine. It can be on your computer or on a server. People install it on servers to have it running 24/7, and that’s really what you want, because keeping your computer plugged in all day might not be the best idea.
I already made a video explaining how to install it on a free Oracle server, but it was a bit technical. Many people wrote to me in the comments, even privately, saying they needed help and didn’t quite know how to do it.
I honestly believe that knowledge should be accessible to everyone. I want everyone to try these tools without needing a technical background.
And it so happens that Hostinger reached out to me and showed me a one-click OpenClaw installer. So I saw a perfect opportunity for all those people who don’t have much technical knowledge and want to install it quickly.
Step 1: Buy the server on Hostinger
On Hostinger’s website there’s a section exclusively dedicated to OpenClaw. From there you can purchase your preconfigured server directly.
There are several options to select the server type. In my opinion, the KVM 2 is more than enough. No need to select a more expensive one, because what this one has is plenty to run OpenClaw:
- 8 GB of RAM
- 2 CPU cores
- 100 GB of NVMe storage (the super fast drives)
With this we can run multiple bots and automations without our server breaking a sweat. And if in the future we need more power, scaling it up is super easy.
Pro Tip: The price is very competitive compared to what’s out there. The more months you select, the cheaper it gets.
What is Nexo and the AI credits?
When selecting the server, you’ll see an option for OpenClaw Nexo credits for $5. They’re not charging you for installing OpenClaw - they’re giving you Nexo credits.
What is Nexo? It’s a unified gateway managed by Hostinger that lets you connect to different AIs (OpenAI, Claude, and a bunch of models) in a unified way.
- If you don’t know much about APIs, don’t deal with models, haven’t spent money on OpenAI or Claude → I recommend selecting it. It’ll come pre-installed and you’ll have $5 in credits, like paying $5 at any provider’s API.
- If you already know your stuff, have your API keys and have already invested → you don’t need to select it.
🎁 Exclusive offer: To get your server on Hostinger, go to hostinger.es/devknives and use code DEVKNIVES for a 10% discount.
Step 2: Configure the server
Once paid, we’re in Hostinger’s interface. Click Setup and it asks us the following:
Location
Always select the one closest to your country. In my case France, which has the least ping. Less ping, more speed.
Operating system
We’ll select Ubuntu and for the version I’ll go with 24.04 LTS, which currently has the most long-term support.
Password and SSH key
It asks us to create a password and an SSH key. This is to access the server from the terminal.
The password is straightforward: set a password that’s long, with special characters, and keep it safe.
Setting up SSH (optional)
To set up SSH, which is optional but more secure, open a terminal and generate the key:
ssh-keygen -t ed25519 -C "youremail@example.com"It will ask for a location. You can leave the default or choose a custom folder. Once generated, display the public key with:
cat ~/.ssh/id_ed25519.pubCopy the content, paste it in Hostinger, and done. SSH key is configured.
Server extras
In the next step, two extras appear:
- Malware scanner: It’s free, I recommend selecting it.
- Docker Manager: I recommend selecting it because we need it to install OpenClaw. But if you’re absent-minded like me and didn’t select it, no worries, I’ll also explain how to do it later.
Wait for everything to configure and done, our server is set up.
Step 3: Connect to the server (optional)
This step is optional, but it can be useful for configuring anything in the future.
Via terminal with password
ssh root@YOUR_SERVER_IPIt will ask for the password you set earlier. Enter it and you’re in.
Via terminal with SSH key
ssh -i /path/to/your/key root@YOUR_SERVER_IPThis lets you in directly without entering a password. It’s the most secure way to access servers.
From the browser (the easiest)
If the terminal isn’t your thing, from Hostinger’s interface there’s an option to access the terminal from your browser. Click the button and it opens a new window with a terminal. No complications.
Step 4: Install Docker Manager
On the Hostinger server there’s a Docker Manager section. If you selected it during initial setup, you already have it. If not, simply go to the section and click Install. That’s it, that’s the only difference.
Wait for it to install and move to the next step.
Step 5: Install OpenClaw with one click
Now the good part. Go to the Catalog section inside Docker Manager. They have a ton of applications available, but we’re interested in OpenClaw. Search for it, click Install and… that’s literally it. Seriously, nothing more to it.
Configure environment variables
Now it shows a window with environment variables, which is basically what OpenClaw needs to work.
First, it lets you add provider APIs like Anthropic, OpenAI, Gemini, etc. If you already have one, go ahead and add it. In my case I used Gemini because it’s free and it worked as an example. It also asks for your phone number with the country code - that’s for WhatsApp configuration.
And then there’s the access token, and pay attention here: copy it and save it somewhere safe. This token is what lets us access our panel and keeps anyone else from getting in. Don’t lose it.
Get a free Gemini API Key
If you don’t have an API key from any provider, no worries - you can grab one from Gemini for free. Just head over to Google AI Studio, select or create a project, generate your API key and paste it into the environment variables. Takes less than a minute.
Nexo (if you didn’t add an API)
If you didn’t enter any API key, the next step also offers you Nexo. You can load whatever balance you want and it directly configures the AI connection for you. If you already added your API key like me, you don’t need this step.
Click Continue, let it install, and simply wait.
Step 6: Access OpenClaw
Once installed, we can expand the arrow and the ports where it’s running appear. Click on the ports, the OpenClaw login page opens, paste the token we copied earlier and… we’re in. Just like that.
Double security verification
There’s one more security step. Go to Overview and paste the same token you pasted before. It’s a double check. Paste it, click Connect, and everything appears.
Fixing a common bug
I ran into a bug that I want to mention in case it happens to you: I was talking to the assistant and it wasn’t responding at all. OpenClaw is still very new and has bugs.
The fix was going to the Models section, opening the dropdown, selecting a different model, and saving. After that, I went back to the Assistant and it was responding perfectly. I assume it’s a configuration issue or some minor bug.
Set up your assistant
With this we’re ready to talk to it. This step is very important and very simple - the hard part is already over.
The assistant will ask about its name, its personality, what its “soul” will be. And here’s my advice: don’t hold back. Write everything about yourself - what you work with, what interests you, what you need day to day. The more you tell it, the better it’ll know you, because it stores everything in its memory and then assists you in a much more personalized way. Don’t just jot down four quick things, because trust me, the difference is massive when you take some time to set it up properly from the start.
Connect with WhatsApp and Telegram
OpenClaw can connect with WhatsApp and Telegram.
WhatsApp (currently has a bug)
To connect it with WhatsApp, go to Channels and click generate QR. In theory you scan it from WhatsApp’s “Link devices”. But as of this post there’s a bug in OpenClaw that makes WhatsApp not work. I’ve seen open issues from people with the same problem. It may be fixed by the time you try it.
Telegram (recommended)
Instead, I recommend connecting via Telegram, which is way easier. Basically, go to Telegram and search for @BotFather (the official bot for creating bots), create yours and it gives you an API key. Pass that key to the OpenClaw assistant and it configures everything on its own. Then just talk to the bot on Telegram, it’ll give you an authorization code, pass it to the assistant, and that’s it - you’re connected.
The best part is that OpenClaw itself walks you through the whole thing. You don’t have to mess with any weird configuration or commands, it handles everything for you.
My favorite automation: Twitter + AI + Notion
And now what I promised: the automation I’m running that’s been saving me a ton of time.
The problem
I have a problem, and I’m sure it happens to you too. On Twitter, TikTok, and many other places, I keep saving stuff that interests me, things I want to check out later. Whether it’s programming, general interest, whatever. I save it and then never look at it. I’ve got a bunch of abandoned saved items, and they’re things that genuinely interested me.
The solution
What I did was teach OpenClaw to connect to the Twitter API (I gave it my authorization tokens so it can read my saved stuff) and also gave it access to my Notion, where I keep all my notes and write everything.
Then, just with words, no configuration whatsoever, I asked it to every morning go through the things I’ve saved, sort them from oldest to newest, consider the priority based on what it knows about me (because it knows my interests), and remind me via Telegram.
How it works day to day
Every morning it tells me: “Hey Edu, here are some things you saved on Twitter. Want to manage them?”
I can dig deeper with it, ask it to search for information on a topic, have it break it down for me. And when I have what I want, I tell it to create a page in Notion with everything we’ve covered.
The result is that saved items don’t pile up anymore. I process them every day, generate useful and well-developed info in Notion, and once I’ve reviewed something, it gets deleted from Twitter. It feeds me content for videos, helps me apply things I’ve learned, and most importantly, I’m actually using my Twitter saves instead of letting them rot there forgotten.
Conclusion
And that’s it. You’ve got your OpenClaw up and running, all quick, with clicks, no programming, nothing complicated.
If you want me to go into more detail on how to build automations from start to finish, let me know in the comments and we can put together a dedicated post for that.
And as always, if you have any problems setting it up, write them in the comments. I read them, I reply, and I try to help whenever I can.
And remember: if you want your server on Hostinger, head to hostinger.es/devknives and use code DEVKNIVES to get a 10% discount.
See you in the next post!
What do you think?
Leave your opinion, question or suggestion. Comments are synced with GitHub Discussions .