Engineer-owned racks · Pacific Northwest · AS13725
AN AI ASSISTANT THAT ACTS, ON A 99.9% SLA
SELF-HOSTED OPENCLAW
OpenClaw is the MIT-licensed, self-hosted AI assistant gateway that answers where you already chat: Telegram, WhatsApp, Discord, Slack, Signal, and more. It reads, replies, browses, runs tools, and schedules work around the clock. That kind of agent belongs on an always-on server of its own. A NoBull Networks VPS is that: AMD EPYC™ cores capped at 60% load, a credit-backed 99.9% network SLA, and racks we own in the Pacific Northwest.
See plans How deployment works
Illustration of chat messages flowing into an OpenClaw gateway server on a NoBull Networks VPS and out to AI model providers

WHY OPENCLAW BELONGS ON A VPS

An assistant that can act for you needs three things your laptop cannot promise: it never sleeps, it stays in its lane, and it answers fast.

Always Actually On

Laptops sleep, assistants should not
  • OpenClaw runs as a daemon and keeps working while you are offline, so its home needs real uptime
  • Credit-backed 99.9% network uptime SLA with redundant switching behind every node
  • Scheduled tasks, reminders, and long-running jobs keep firing at 3am without your help

A Clean Blast Radius

Give the agent its own box
  • OpenClaw can execute commands and manage files, which is the point, and also the reason it should not live on your personal machine
  • A dedicated VPS is a crisp isolation boundary with full root for you and nothing else on it
  • Pairs with OpenClaw's own tokens, allowlists, and safety controls

Snappy Where It Counts

4.5 GHz cores, quick replies
  • The gateway is a Node.js process, so 4.5 GHz single-core clocks keep every reply quick
  • Direct Tier-1 peering out of Seattle keeps round trips to model APIs short
  • TB-scale monthly transfer covers browsing, media, and busy group chats without rationing
EVERY CHAT, ONE GATEWAY

The Gateway is OpenClaw's single source of truth for sessions, routing, and channel connections. It runs on your VPS, and only the model calls leave the box.

Personal chats Telegram, WhatsApp, Signal Team chats Discord, Slack, Teams More channels Matrix, Google Chat, plugins YOUR NOBULL NETWORKS VPS OpenClaw Gateway sessions, routing, channels Workspace and skills memory, files, tools on NVMe You, privately dashboard over SSH or tailnet Frontier models your Anthropic or OpenAI keys Local models Ollama, on larger plans
OpenClaw is developed in the open by the OpenClaw Foundation, a nonprofit, under the MIT license. It was formerly known as Clawdbot and then Moltbot, so guides under those names still apply.
RECOMMENDED PLANS FOR OPENCLAW

The Gateway itself is lightweight; headroom is for what you ask it to do. Every plan is online in under 3 minutes with full root access, and the benchmarks behind it are published.

Performance VPS · P20Solo assistant
  • 2 vCPU
  • 4 GB RAM
  • 50 GB RAID NVMe Storage
  • 8 TB Monthly Bandwidth
  • Comfortable for a personal agent
$16.00/mo
Get this server
Recommended
Performance VPS · P40The daily driver
  • 2 vCPU
  • 8 GB RAM
  • 100 GB RAID NVMe Storage
  • 16 TB Monthly Bandwidth
  • Headroom for browsing and busy skills
$32.00/mo
Get this server
Performance VPS · P50Heavy and multi-agent use
  • 3 vCPU
  • 12 GB RAM
  • 150 GB RAID NVMe Storage
  • 24 TB Monthly Bandwidth
  • Room for several workspaces at once
$48.00/mo
Get this server
Model usage is billed by your AI provider, not by us. Your NoBull Networks invoice stays a flat server price no matter how much your assistant talks. See the receipts for the benchmarks behind every plan.
Powerful Tools, Held Properly

Isolation Is A Feature, Not Paranoia

We will be straight with you: an autonomous agent that reads messages and runs commands deserves respect. OpenClaw ships pairing tokens, allowlists, and safety controls, and its docs recommend keeping the dashboard private behind SSH or a tailnet. Use them.

What we add is the part you cannot configure in software: a dedicated, isolated server with nothing else on it, full root for you, and engineers keeping the infrastructure honest underneath, 24/7.

This is an unmanaged deployment. You own the agent and its guardrails; we own the uptime.

  • MIT-Licensed Open Source developed by the OpenClaw Foundation
  • Dedicated Isolation Boundary the agent gets its own box, not your laptop
  • Always-On Daemon Home 99.9% network uptime SLA, credit-backed
  • 4.5 GHz Single-Core Clocks benchmarks published on VPSBenchmarks
  • Dedicated IPv4 Included stable pairing for every channel
  • Full Root Access your keys, your rules, your kill switch
FROM ORDER TO FIRST REPLY

OpenClaw's onboarding wizard does most of the work. Bring a model API key and one chat app to pair.

Order your VPS

Pick a plan and check out. Your server is online in under 3 minutes with full root access and a dedicated IPv4 address.

Install Node and OpenClaw

OpenClaw wants a current Node.js (22.22.3 or newer; 26 recommended). Then npm install -g openclaw and run the onboarding wizard with its daemon install.

Pair your channels

Connect Telegram, WhatsApp, Discord, Slack, or any supported channel, and add the API key for your model provider.

Lock it down, let it work

Set allowlists for who may talk to it, keep the dashboard behind SSH or your tailnet, and hand your assistant its first standing job.

Full installation, channel, and security documentation lives at docs.openclaw.ai, maintained by the OpenClaw community.
Frequently Asked Questions
Self-hosting OpenClaw on a NoBull Networks VPS

Explore the knowledge base or open a ticket in the Console. An engineer answers.

What exactly is OpenClaw?
OpenClaw is an open source, MIT-licensed gateway that connects your messaging apps to AI agents that can actually do things: reply in chat, browse, manage files, run tools, and schedule tasks. It is developed by the OpenClaw Foundation, a nonprofit, and was formerly known as Clawdbot and later Moltbot.
Why run it on a VPS instead of my own computer?
Two reasons. Availability: an assistant that schedules and acts for you is only useful if it is awake, and a VPS with a 99.9% network uptime SLA beats a laptop lid. Isolation: OpenClaw can execute commands, so giving it a dedicated server keeps a clean boundary between your agent and the machine your tax returns live on.
What do I need to get started?
Three things: a VPS (the P20 is a comfortable start, the P40 gives headroom), an API key with an AI provider such as Anthropic or OpenAI with billing enabled, and one chat app to pair. OpenClaw requires a current Node.js release: 22.22.3 or newer, with Node 26 recommended.
Which chat apps does it support?
The docs list Telegram, WhatsApp, Discord, Slack, Signal, Matrix, Microsoft Teams, Google Chat, Zalo, and more, with additional channels like Twitch available as plugins. You can connect several at once; the Gateway routes them all.
Is it safe to run an autonomous agent?
It is as safe as you configure it to be, and we would rather say that than pretend otherwise. OpenClaw ships pairing tokens, allowlists, and safety controls, and recommends keeping its dashboard private behind SSH or a tailnet. A dedicated VPS adds the structural piece: whatever the agent does, it does inside its own box.
What does it cost to run?
OpenClaw itself is free. You pay for the VPS at our usual flat pricing, and your AI provider bills model usage on their side by tokens. Busy assistants cost more in model usage, not in hosting; your NoBull Networks invoice does not change.
READY TO HIRE YOUR OWN AGENT?

Order a VPS, run the onboarding wizard, and your assistant answers its first message today.

Performance VPS
P40 at $32.00/mo Get this server
All VPS Plans
Compare every tier See VPS plans
More Solutions
See what else fits Browse solutions