ReadyOpenClaw — server for an AI agent. YOUR agentON A DEDICATED SERVER.
We provision, harden, and run the VPS. You pick the runtime — OpenClaw, Hermes, or Codex — and configure the agent for your work.
01 — Three runtimes
Three runtimes
OpenClaw, Hermes Agent, and Codex run on the same server and plan. Claude Code and Vercel Eve are coming soon, with no date. More: OpenClaw and Hermes Agent. Compare them in the questions on the OpenClaw page.
02 — Launch
How to deploy
The live first-run path: create an agent, connect a model through Codex Auth, deploy, and link Telegram. No SSH and no config files. All of it in the console.
First launch · no card
24 hours on a trial VPS, then the plan you pick
03 — Codex
Codex on this server
Codex is the third live runtime, not a separate landing. You pick it when you create an agent: your own key or Codex Auth. The model is not in the plan.
04 — What is included
The agent is yours. The server is ours.
One price covers the server and its upkeep. You pay the model provider directly. We do not mark that up.
What is not included
Exactly one line: model cost. Everything else, including upkeep, is in the plan.
- vpsDedicated VPS2 vCPU · 4 GB · 40 GBA private virtual server for your agent. Capacity is not shared with other customers, and your data does not sit next to theirs.
- channelsMessaging channelsTelegram, WhatsApp
Telegram and WhatsApp connect in guided steps, with a live check. Discord and Slack follow as the platform is ready. - modelThe model you chooseYour key or Codex Auth
Your provider API key, or sign-in through Codex Auth. Keys stay encrypted and are passed only to the agent you picked. - upkeepUpdates and watchMonitoring and recoveryWe watch the server, update the platform, and restore the agent after an infrastructure failure.
- consoleEnglish consoleNo SSH, no config filesSetup, agent status, logs, and billing live in one interface. No SSH and no hand-edited configs.
- billingModel costNot included · billed by the providerToken usage is billed by the model provider on your key. We do not mark it up.
05 — Control
Your secrets are not our product
Secrets stay encrypted. An agent only receives the values its current settings need.
Keys stay encryptedChannel tokens and model keys are passed to a given agent only when its settings are applied.
You pick the modelYour own provider key or Codex Auth. The platform does not sit between you and the model.
Visible isolationA dedicated VPS instead of a shared box. You can see where the agent runs, and that nobody else is on that server.
( values never leave the vault in the clear )
06 — Plans
Plan and trial
Model cost is not in the plan: you pay the provider directly. The first 24 hours need no card. Annual billing includes two months free.
Loading the USD catalogue.
07 — FAQ
The short version
Which runtimes can I run?
Today you can run OpenClaw, Hermes, and Codex. The runtime is chosen when you create an agent. The server stays the same. Claude Code and Vercel Eve are on the roadmap, with no promised date.
Do I need to administer a server?
No. We create the VPS, deploy the agent, and run the infrastructure. You configure the agent and connect the channels you want.
Who pays for the model?
You do, directly to the provider, with your own key or through Codex Auth. The plan covers the server and its upkeep.
Can I cancel?
Yes. Turn off renewal in the console before the next billing period.
First 24 hours free
Deploy an agent today
A trial VPS with no card. Tell us what the agent is for and we will help you pick a plan and connect channels.