IR
IT-Roc
Platform & Store
Login
Home›Servers›AI Agent Server
🤖 AI Agent Server

Your AI agent keeps working, even while you sleep.

A server of your own with a preinstalled agent framework, where your AI agents run permanently. Connect it to your messengers and choose the language model with your own API key.

🤖
Servers
AI Agent Server
…
  • ✓ Agent framework preinstalled
  • ✓ Connects to Telegram, WhatsApp or Slack
  • ✓ Free choice of model with your own API key
Configure now

All prices exclude VAT.

🇩🇪

Data centres in Germany

Operated in ISO 27001-certified data centres, GDPR-compliant.

🔐

Secure by default

Hardened defaults, regular patching and automatic backups.

⏱️

24/7 monitoring & SLA

Monitored around the clock, incident response under the agreed SLA.

🤝

Personal support

Dedicated contacts at IT-Roc instead of an anonymous call centre.

✦ Your benefits

Why customers choose IT-Roc.

1

Always available

Your agent runs on a server, not on your laptop. It keeps responding and working even when every computer is switched off.

2

Free choice of model

Add your own API key and use the model that fits your task. Model usage is billed directly by the provider.

3

Where you already chat

The agent is reachable via Telegram, WhatsApp or Slack. You do not need a new app to work with it.

4

Your own environment

Configuration, history and credentials live on your own server in a German data center, separate from other customers.

✦ Pricing

Transparent plans, no hidden costs.

Cancel monthly or save with annual billing. Upgrade to a bigger plan at any time.

Loading plans …
✦ How it works

Up and running in four steps.

  1. 01

    Choose a plan

    S or M is enough for one agent. For several agents or local models, go for L.

  2. 02

    Server provisioning

    We set up the server with the agent framework and send you your login details.

  3. 03

    Connect model and channel

    Add your API key and connect the agent to Telegram, WhatsApp or Slack.

  4. 04

    Put your agent to work

    Give the agent tasks and tools. From now on, it runs permanently on your server.

✦ Who it's for

Made for your project.

An agent that takes care of appointments, notes and research for you. Just send it a message in your messenger.

Get your own AI agent.

Choose the right plan and your agent will be ready to work in no time.

Talk to an expert
✦ Technical details

Everything under the hood.

  • › S: 2 vCPUs, 4 GB RAM, 80 GB NVMe
  • › M: 4 vCPUs, 8 GB RAM, 160 GB NVMe
  • › L: 8 vCPUs, 16 GB RAM, 320 GB NVMe
  • › Several agents in parallel in plan L
✦ FAQ

Frequently asked questions.

Which agent framework is installed?+

The server comes with a common agent framework preinstalled, for example OpenClaw or Hermes. We are happy to discuss which one best suits your project.

Are the costs for the AI model included?+

No. You use your own API key and pay for model usage directly to the respective provider. In return, you are completely free in your choice of model.

Can I run a model locally?+

Plan L supports local models with up to around 8B parameters. For larger models, we recommend connecting via API.

Which channels can the agent be reached on?+

The agent can be connected to Telegram, WhatsApp or Slack. That way you use it where you already communicate.

Can I run several agents?+

In plan L, several agents run in parallel. In S and M, the server is designed for one agent.

✦ Personal consultation

Questions about AI Agent Server?

Our experts will advise you personally and work out the right solution with you. Fill in the form and we will get back to you shortly.

  • ✓ Free and without obligation
  • ✓ Reply within one business day
  • ✓ Direct contact with our engineers

By submitting you consent to us processing your details to handle your request. See our privacy policy.

✦ Goes well with

Other services that combine well.