Skip to content
Sign up Log in
One-click app

Open WebUI & Ollama

Your own AI chat in the browser: Open WebUI with Ollama runs on your server, Llama 3.2 is preinstalled, and you add more open models with one click.

Choose Package Configure Server

Monthly cancellable · no setup fee · prepaid

Recommended spec

The optimal configuration

Our recommendation for Open WebUI + Ollama: the cheapest package with the right resources – no minimum term.

4
CPU Cores
AMD EPYC vCores
16GB
Memory
DDR4 ECC
80GB
NVMe Storage
NVMe SSD · Ceph HA
9,99€
Price
Monthly

Minimum requirements of Open WebUI + Ollama (below them the panel refuses the install): 8 GB RAM · 50 GB NVMe

In Detail

Why an AI Chat on Your Own Server?

Prompts and documents stay on your server instead of going to an external AI service. There is no billing per token or user – you pay for the server, the open models are free.

Open WebUI is a ChatGPT-style chat interface, Ollama runs the language models. The panel installs both together and pulls Llama 3.2 (3B) along, so you can chat right after the install. You add more models such as Mistral, Qwen or Gemma in the interface. The panel creates your admin account with your e-mail address; new sign-ups need your approval. To set expectations: our root servers have no graphics card, the models run on the CPU cores. Small models like Llama 3.2 respond at a usable pace, large models get noticeably slower. That is why the app requires at least 8 GB RAM and 50 GB storage.

  • DDoS Protection
  • Location — Frankfurt DE-CIX
  • Storage — NVMe Ceph HA

Features

All Benefits at a Glance

Full root access — install anything

Install any software, any runtime, any version — no restrictions from panels or shared hosting.

NVMe SSD Ceph HA storage

High-performance NVMe SSDs in a Ceph HA cluster for maximum I/O performance and 3x redundancy.

Full SSH root access

Connect anytime via SSH and manage your server with full administrator privileges.

Enterprise DDoS protection included

Permanent protection with high-capacity filtering — included free with every server.

Runs continuously — no sleep mode

No sleep mode and no shutdown when idle: your server keeps running as long as its term is paid.

Low latency — Frankfurt DE-CIX

Located directly at DE-CIX in Frankfurt for low latency to European networks.

Always included

Included with every root server

Enterprise DDoS protection included Full root access — install anything NVMe SSD Ceph HA storage Runs continuously — no sleep mode Low latency — Frankfurt DE-CIX 1× IPv4 + /64 IPv6

All Root Server Packages

Choose the package that best fits your project

Every one of these packages meets the minimum requirements of Open WebUI + Ollama.

RootPaket-3

Recommended

9,99€ Monthly

  • 4 AMD EPYC vCores
  • 16 GB DDR4 RAM
  • 80 GB NVMe SSD
  • 1× IPv4 + /64 IPv6
  • DDoS Protection Included
View Packages

RootPaket-4

19,69€ Monthly

  • 6 AMD EPYC vCores
  • 24 GB DDR4 RAM
  • 130 GB NVMe SSD
  • 1× IPv4 + /64 IPv6
  • DDoS Protection Included
View Packages

RootPaket-5

30,49€ Monthly

  • 8 AMD EPYC vCores
  • 32 GB DDR4 RAM
  • 180 GB NVMe SSD
  • 1× IPv4 + /64 IPv6
  • DDoS Protection Included
View Packages

RootPaket-6

46,79€ Monthly

  • 8 AMD EPYC vCores
  • 48 GB DDR4 RAM
  • 200 GB NVMe SSD
  • 1× IPv4 + /64 IPv6
  • DDoS Protection Included
View Packages

Get Started in 3 Steps

Get Started in 3 Steps

1

Order a root server with at least 8 GB RAM

2

Pick Open WebUI under “Apps” in the panel, optionally enter a domain

3

Log in and start chatting with Llama 3.2

  • Admin account created during the install
  • Own domain with HTTPS (Let’s Encrypt) possible
  • The install usually takes about 8 minutes
  • Runs on Ubuntu, Debian, AlmaLinux and Rocky Linux

More one-click apps: WordPress · Nextcloud · Vaultwarden · Coolify · Jellyfin · Forgejo / Gitea · Grafana · Nginx Proxy Manager · Paperless-ngx · Minecraft server · Docker / Portainer · WireGuard (wg-easy) · n8n

Frequently Asked Questions

Answers to the most important questions

Does it run without a graphics card?
Yes. Ollama computes on the CPU cores. Small models with a few billion parameters are usable that way; large models respond much more slowly than services with a GPU. More vCores and RAM help noticeably.
Which models can I use?
Llama 3.2 (3B) is preinstalled. You add more open models from the Ollama library – such as Mistral, Qwen or Gemma – in Open WebUI. Make sure the model fits into your package’s memory.
Can several people use the chat?
Yes. New accounts first land in a queue and only become active once you approve them as admin. That keeps the chat private even if the address is publicly reachable.
Can I connect external AI services too?
Yes. Besides Ollama, Open WebUI supports OpenAI-compatible APIs. Requests then go to that service, though – only the models running via Ollama stay local.

Get Started Now

Your own AI chat without per-token billing – from €9.99 a month.

Choose Package Configure Server