Run OpenClaw on a VPS
A VPS is the no-hardware way to keep an OpenClaw Gateway online. It is a Linux box in a data center that stays up around the clock, answers from anywhere, and disappears the moment you delete it. Onepilot deploys the Gateway to it over SSH and puts the whole cockpit on your iPhone.
TL;DR
The OpenClaw Gateway runs comfortably on a small Linux VPS.With a hosted LLM provider doing the model work, 1 vCPU and 1 GB of RAM is enough to run the Gateway process, its memory, and a messaging channel. Spin up Ubuntu 24.04 at any host you like. Onepilot opens an SSH session, installs the OpenClaw CLI under an nvm prefix with no sudo, starts the Gateway as a background process, and lets you pick the LLM provider and the channel it answers on (Telegram, Discord, or Slack) inside the app. The same SSH tunnel gives you the server's terminal, file browser, git, and cron on the phone, so nothing new has to listen on a public port.
Run the OpenClaw Gateway on a VPS, reachable from anywhere, driven from your iPhone. Download it on the App Store.
What size VPS does OpenClaw need?
OpenClaw runs on a small instance. With a cloud LLM provider, the Gateway process, its memory, and a messaging channel fit inside 1 GB of RAM on 1 vCPU. The heavy compute happens at the provider, not on the box, so the server only carries the agent loop and its connections.
Move up to 2 GB if you run several channels at once, do heavy browser automation, or want headroom for bursts. You only need a large instance if you plan to run a local model on the server itself, which is rarely worth it on shared CPU hardware. Point OpenClaw at a provider instead and keep the VPS small.
Which VPS provider should I pick for OpenClaw?
Any Linux host works, because Onepilot only needs a box that runs Linux and accepts SSH. The major budget providers all offer an instance in the 1 GB to 2 GB class, including ARM options and free tiers. Hetzner, DigitalOcean, AWS Lightsail, and Oracle Cloud Free Tier are all common picks.
Choose on region, bandwidth, and reliability rather than brand. The OpenClaw Gateway behaves identically regardless of whose hardware it sits on, so the practical question is latency to you and to your LLM provider, not the logo.
Is it safe to run OpenClaw on a public VPS?
Yes, with standard hygiene. A VPS has a public IP, so the security model matters more than it does on a home machine. Use SSH key authentication only, disable password login, run a firewall that exposes nothing but the SSH port, and turn on unattended security updates. That posture is boring and effective.
The biggest risk people add themselves is exposing a browser control panel on the public IP, since that is an extra web service that can drive a shell-capable agent. Onepilot avoids that surface entirely. It reaches the Gateway over the same SSH you already lock down and gives you the terminal, file browser, git diffs, and cron on the phone, so nothing new listens on a public port.
Disposable by design
A VPS is cattle, not a pet. Onepilot installs OpenClaw under nvm in your user directory and keeps it reversible, so rm -rf ~/.nvm ~/.openclaw removes the agent and destroying the instance removes everything. Back up the OpenClaw config and memory before big changes, and you can rebuild from scratch in minutes.
FAQ
What are the VPS requirements for OpenClaw?
OpenClaw is light on a VPS. With a hosted LLM provider, the Gateway runs comfortably on 1 vCPU and 1 GB of RAM, because the model runs at the provider and the server only carries the agent's control loop, its memory, and the messaging channel. Move to 2 GB if you run multiple channels or heavy browser automation. You only need a large instance if you intend to run a local model on the server, which usually is not worthwhile on shared-CPU VPS hardware.
Which VPS provider is best for OpenClaw?
Any Linux host works, since Onepilot is provider-agnostic and only needs a box that runs Linux and accepts SSH. The major budget providers all offer an instance in the 1 GB to 2 GB class, including ARM options and free tiers such as Oracle Cloud Free Tier. Pick on region, bandwidth, and reliability. The Gateway behaves the same regardless of whose hardware it runs on.
Is it safe to run the OpenClaw Gateway on a public VPS?
Yes, with standard hygiene: SSH key authentication only, no password login, a firewall that exposes only the SSH port, and automatic security updates. The largest risk people add is exposing a browser control panel on the public IP, because that is an extra web service controlling a shell-capable agent. Onepilot avoids that, since it reaches the Gateway over SSH and adds no public surface.
VPS or Raspberry Pi for OpenClaw?
A VPS has no hardware to buy, is reachable from anywhere by default, and offers better uptime and bandwidth, in exchange for an ongoing host bill. A Raspberry Pi is hardware you own with only an electricity cost and full physical control, which suits a home setup. Onepilot deploys OpenClaw the same way to both, so the choice comes down to whether you want a box in a data center or one at home.
How do I supervise the VPS Gateway from my phone?
Onepilot connects to the VPS over SSH and gives you the server's real terminal, a syntax-highlighted file browser to read the skill files and config OpenClaw writes, a git tab with real diffs, and cron, all on the iPhone. You direct the agent through its channel (Telegram, Discord, or Slack) and use the app to inspect, fix, and schedule. The same app also supervises Hermes and runs Claude Code or Codex on the same VPS.
Run the OpenClaw Gateway on a VPS from your phone
Download Onepilot on the App Store.
See also: Install OpenClaw on any server, OpenClaw setup & keep-alive, What is OpenClaw?, Run OpenClaw on iPhone.