OpenClaw hosting
Your OpenClaw agent,
always on, set up in one click.
Run OpenClaw on a VPS that never sleeps, so your AI agent answers on WhatsApp and Telegram day and night. Pick OpenClaw when you create a server and we install it, keep its dashboard private behind an access token, and hand you a one-click sign-in. No Node.js, no commands.
- Private by default
- No commands
- Runs 24/7
- From ₹548/month
Step 1 of 5 · Pick the app
What you'd normally do
curl -fsSL deb.nodesource.com/setup_24.x | bashInstall Node.js 24npm install -g openclawInstall OpenClawuseradd -m openclaw && loginctl enable-lingerA separate user for the agentopenclaw onboard --install-daemonGateway, token and servicessh -N -L 18789:127.0.0.1:18789 …Tunnel in to reach the dashboardapt install caddyHTTPS, if you want it public
The Iotamine control panel, step by step. Timings are from a real install: server online in 29 seconds.
At a glance
What you get with an OpenClaw VPS on Iotamine
- What it is
- A cloud VPS with OpenClaw pre-installed and running. It is self-hosted OpenClaw on a server that is only yours, online 24/7.
- What is installed
- The official OpenClaw package on Node.js 24, on Ubuntu 24.04. It runs as its own non-root user and as a service that starts again after a reboot.
- Dashboard access
- Private by default: reachable only through an SSH tunnel, with your access token, as OpenClaw recommends.
- Publishing
- One click in the control panel puts the dashboard on HTTPS at <your-ip>.ip.iotamine.cloud, still protected by your token and OpenClaw's device approval. One more click makes it private again.
- Sign-in
- Open dashboard in the control panel gets a one-time sign-in link from your server in about 10 seconds. The access token is on the server's page and can be rotated any time.
- AI models and chat apps
- Bring your own AI model key and connect WhatsApp, Telegram and other chat apps from OpenClaw's own dashboard.
- Setup time
- The server is online in under a minute, and OpenClaw starts right after. We email you when it is ready.
- Sizes and price
- From 1 vCPU and 2 GB RAM at ₹548 a month. We recommend 2 vCPU, 4 GB RAM and 40 GB disk: ₹1,015 a month (₹1.41/hour).
- Location
- New Delhi, India. Your agent's memory, files and settings stay on your server, in India.
- Access
- Full root SSH access. It is a normal VPS, so you can add tools for your agent or run other apps next to it.
- Billing
- Hourly, in rupees, with UPI. Resize CPU and RAM separately; delete the server and billing stops.
- Updates
- You decide when to update OpenClaw, as with any self-hosted install. Take a snapshot first so you can roll back.
How to self-host OpenClaw
Five steps, all clicks
- 1
Choose OpenClaw. On the deploy page, open the Apps tab and pick OpenClaw. The recommended size is filled in for you.
- 2
Review and deploy. Check the location, size and hourly price, then click Deploy server.
- 3
Wait about a minute. The server builds in under a minute and OpenClaw starts right after. We email you when it's ready.
- 4
Open the dashboard. Click Open dashboard for a one-time sign-in link, or keep it private and use the SSH tunnel shown on the server's page.
- 5
Add a model and chat apps. In OpenClaw, add your AI model key and connect WhatsApp, Telegram or another chat app. Then just message your agent.
Included
Set up the way OpenClaw recommends
An agent that can read your messages and use tools should not have an open dashboard. Ours starts locked down, and you open it up only if you choose to.
Private by default
The dashboard listens only on the server itself. Nothing is reachable from the internet until you decide otherwise.
An access token you control
A strong token is generated for you and shown on the server's page. Rotate it with one click if it ever leaks.
Publish over HTTPS in one click
Want it in your browser without a tunnel? Publish it on a ready HTTPS address, still behind your token and device approval.
One-time sign-in links
Open dashboard asks your server for a fresh sign-in link, so you never paste the token into a browser.
Its own user, always running
OpenClaw runs as a separate non-root user and as a service, so it keeps going after reboots and stays away from root.
Snapshots and resizing
Take a snapshot before you try a new plugin, and grow CPU, RAM or disk on their own as your agent does more.
Ideas
What people use OpenClaw for
You message your agent like a person, from the chat app you already use. A few common starting points:
Inbox summaries
“Every morning, summarise my unread email and draft replies.”
TelegramReminders and to-dos
“Remind me to renew the domain on Friday at 10.”
WhatsAppResearch
“Find the three best-reviewed tools for this and compare their prices.”
TelegramDaily briefings
“At 9 every weekday, send me the news on our industry in five bullet points.”
WhatsAppTeam helper
“Answer the team's questions about our docs in the support channel.”
DiscordFiles and notes
“Save this to my notes and pull up what I wrote about it last month.”
TelegramPricing
OpenClaw VPS hosting prices, billed by the hour
You pay for the server only. OpenClaw itself has no extra charge from us; your AI model usage is billed by your model provider. Pick any size and resize later.
Starter
₹548/month
₹0.76/hour · 5 TB bandwidth
Try OpenClaw out with one chat app and a few light tasks.
Standard
Recommended₹1,015/month
₹1.41/hour · 8 TB bandwidth
The size we recommend. A personal agent on several chat apps, running all day, with room to grow.
Power
₹1,931/month
₹2.68/hour · 16 TB bandwidth
Several agents, many skills and plugins, or browser and file tasks running at once.
Team
₹3,761/month
₹5.22/hour · 32 TB bandwidth
An agent a whole team talks to, with heavy tools and lots of memory.
New Delhi prices including one IPv4 address, for a full 720-hour month. Delete the server and billing stops.
Where to run it
Your own computer vs a VPS
Many people start OpenClaw on a laptop or a spare mini PC at home. It works, until the machine sleeps, the power goes or the agent needs to stay away from your personal files.
| Laptop or home PC | Iotamine OpenClaw VPS | |
|---|---|---|
| Always on | Stops when the computer sleeps, restarts or loses power | Runs 24/7 in a data centre |
| Reachable | Behind your home router; webhooks and remote access need extra set-up | Public IP, plus HTTPS in one click if you want it |
| Your personal files | The agent shares a machine with your own files and accounts | A separate server with only what you give it |
| Set-up | Node.js, the install, the service and the tunnel, by hand | One click, ready in about a minute |
| Dashboard safety | Up to you to keep it off the internet | Private by default, with a token you can rotate |
| Cost | The hardware, plus electricity and your internet connection | From ₹548/month, billed hourly |
Choosing a host
What to check in OpenClaw hosting
| Question to ask | Why it matters | Iotamine |
|---|---|---|
| Is the dashboard private by default? | An open OpenClaw dashboard can give strangers control of an agent that reads your messages. | Yes, SSH tunnel and token only |
| Can you rotate the access token? | If a token leaks you need to cut it off straight away. | One click on the server's page |
| Is there a safe way to use it in a browser? | Tunnels are fiddly on a phone or a borrowed computer. | HTTPS + token + device approval, one click |
| Does it run as root? | An agent with tools should not have full control of the server. | No, its own non-root user |
| Does it stay up after a reboot? | Your chat apps expect the agent to always answer. | Yes, it runs as a service |
| Where is the data stored? | Your agent's memory holds personal details from your chats. | New Delhi, India |
| Can you size it to your workload? | More plugins and parallel tasks need more RAM. | Change CPU, RAM and disk separately |
| How is it billed? | Hourly billing lets you try it without a long commitment. | Hourly, in rupees, UPI accepted |
FAQ
OpenClaw hosting questions
OpenClaw is an open-source personal AI agent. It runs on your own computer or server, connects to chat apps like WhatsApp and Telegram, and uses the AI model you choose to answer you, remember things and carry out tasks with its tools.
Running OpenClaw on a server that is always online instead of your own computer, so your agent answers at any time. On Iotamine you get your own VPS with OpenClaw already installed and running, with a private dashboard, ready for you to add a model and chat apps.
Create an account, open the deploy page, choose the Apps tab and pick OpenClaw, then click Deploy server. The server is online in under a minute and OpenClaw starts right after. Click Open dashboard on the server's page for a one-time sign-in link, then add your AI model key and chat apps in OpenClaw. You don't type any commands.
No. It starts private: it only listens on the server itself, and you reach it through an SSH tunnel with your access token, as OpenClaw recommends. If you want it in a browser, Publish over HTTPS puts it on an HTTPS address that still needs your token and OpenClaw's own device approval. Make private turns that off again.
Click Open dashboard on the server's page in the control panel. It asks your server for a one-time sign-in link, which takes about 10 seconds, and opens the dashboard with it. You can also keep it private and use the SSH tunnel command shown on the same page.
Not to try it, but to keep it. On a laptop the agent stops whenever the computer sleeps, and it shares the machine with your personal files and accounts. A VPS keeps it online 24/7, separate from your own computer, with a public IP for chat apps and webhooks.
OpenClaw needs Linux and Node.js, which we install. It uses about 1 GB of RAM when idle, so 1 vCPU and 2 GB RAM is enough to try it (about ₹548 a month here). For daily use we recommend 2 vCPU, 4 GB RAM and 40 GB of disk. Add more RAM if you run many plugins or several tasks at once.
You pay for the server by the hour. The smallest size is about ₹548 a month in New Delhi; the size we recommend (2 vCPU, 4 GB RAM, 40 GB disk) is ₹1.41 per hour, about ₹1,015 for a full month, including one IPv4 address and 8 TB of bandwidth. There is no extra charge for OpenClaw. Your AI model usage is billed separately by your model provider.
You add them in OpenClaw's own dashboard: bring an API key from the model provider you prefer, and connect chat apps such as WhatsApp and Telegram. Iotamine doesn't limit which ones you use.
It is the official OpenClaw package, self-hosted on your server. Iotamine installs and configures it; we are not affiliated with the OpenClaw project.
You do, as with any self-hosted install. You have full root access, so you can update OpenClaw whenever you choose. Take a snapshot first so you can roll back if something changes.
On your server's disk in New Delhi, India: its memory, settings and files. Your messages go to the AI model provider you connect, under that provider's terms. You can take snapshots from the control panel at any time.
Get started
Your own AI agent, online a minute from now.
₹800 signup bonus after KYC ($10 for international accounts). Then pick OpenClaw on the deploy page.