Give your OpenClaw assistant an always-on home
OpenClaw answers you in WhatsApp, Telegram, Slack and Discord, but only while the machine it runs on is awake. Put it on a RelyHost VPS and it stays reachable around the clock, away from your personal computer.
An open-source AI assistant that lives in your chats
OpenClaw is a free, MIT-licensed AI assistant. You message it from the apps you already use, and it can run commands, manage files, browse the web and use any skills you install.
It works with the model of your choice, such as Claude, OpenAI's models or a local model. Its memory, settings and credentials stay on the server you run it on.
Works where you chat
WhatsApp, Telegram, Slack, Discord, Signal, iMessage and 20+ other channels.
Bring your own model
Connect a hosted or local model, and switch later without starting over.
Open source
Developed in the open by the non-profit OpenClaw Foundation.
Why run OpenClaw on a server
Always reachable
Laptops sleep, travel and restart. A VPS stays online, so your assistant can answer a message at 3am as easily as at noon.
Off your personal machine
OpenClaw runs its tools directly on the host. A separate server keeps that access away from your own files, accounts and browser.
Full root access
Use the official installer, Docker, Nix or npm. It is your server, so any install method and any extra tooling works.
Room to grow
Move to a bigger plan as you add channels, skills and browser automation. Your workspace and memory come with you.
What OpenClaw needs
The model runs at your provider, so the server doesn't need a GPU. Its resources go to the OpenClaw gateway, browser automation and the tools you add. We recommend 2 vCPUs and 4 GB of RAM as a comfortable starting point, and more if your assistant drives a browser often.
Every plan includes NVMe storage and DDoS protection. Ubuntu 24.04 LTS works well with the official installer.
Up and running in three steps
The official installer sets up everything OpenClaw needs, then walks you through onboarding.
Order a VPS
Pick a plan and choose Ubuntu 24.04 LTS. Your server is set up automatically and you get root SSH access.
Run the installer
SSH in and run the official install script. It installs Node.js and OpenClaw, then opens the setup wizard.
Connect a model and your chats
Add an API key for your model provider, then link WhatsApp, Telegram, Slack, Discord or another channel.
Why not just run it at home?
OpenClaw runs fine on a laptop. Here's what changes when it lives on a server.
OpenClaw VPS plans
Every plan includes DDoS protection, NVMe storage and a dedicated IPv4 address on a 1 Gbps uplink.
OpenClaw Starter
- 2 vCPU Cores
- 4 GB RAM
- 60 GB NVMe SSD
- 4 TB Bandwidth
- 1 Dedicated IPv4
Plenty for one assistant with a few channels and skills.
Get StartedOpenClaw Plus
- 4 vCPU Cores
- 8 GB RAM
- 120 GB NVMe SSD
- 6 TB Bandwidth
- 1 Dedicated IPv4
More headroom for frequent browser automation.
Get StartedOpenClaw Pro
- 6 vCPU Cores
- 16 GB RAM
- 180 GB NVMe SSD
- 8 TB Bandwidth
- 1 Dedicated IPv4
For several assistants or heavy tool use.
Get StartedReady to move OpenClaw to a server?
Plans start at $7.50/mo.
Not sure which plan fits, or stuck during setup? Our support team can help you size your server and get your assistant online.