Stable Diffusion Hosting — from €5.49/mo
Generate images on your own server — a VPS for Stable Diffusion and WebUI, 5-hour test period, so you risk nothing
Start generating now
Stable Diffusion hosting is a virtual server with full root access for generating images with a neural network. You install Linux, Python and a WebUI (such as AUTOMATIC1111 or ComfyUI) yourself, upload your own models and checkpoints and work from the browser. The heavy generation runs on the server rather than your PC, and NVMe drives keep large model weights close at hand. Plans start at €5.49/mo, with a 5-hour test period — check the generation speed with your own settings before paying.
Plans
VPS for Stable Diffusion
Standard 1 and Standard 2 in this block are enough to bring up the WebUI and sort out the environment. Practical generation needs a graphics card, and a model library needs 400 GB NVMe or more (Max 1). GPU configurations are built separately. 5-hour test period before payment.
Compare all plansStandard 1
| 4 vCPU · AMD EPYC |
| 4 GB RAM |
| 50 GB NVMe |
| Unlimited traffic |
| 1× IPv4 · KVM |
| 24/7 support |
Standard 2
| 4 vCPU · AMD EPYC |
| 6 GB RAM |
| 60 GB NVMe |
| Unlimited traffic |
| 1× IPv4 · KVM |
| 24/7 support |
Your own WebUI and models
Install AUTOMATIC1111, ComfyUI or another interface, load any checkpoints, LoRA and extensions with no restrictions.
Generation in the cloud
Heavy generation jobs run on the server around the clock without tying up your computer — work from the browser.
Models on NVMe
Fast drives and spare RAM keep large weights, checkpoints and results right next to the generation.
How to run Stable Diffusion on a VPS
Choose a plan and an image
Pick a plan — Standard 1 (4 vCPU, 4 GB) or Standard 2 (4 vCPU, 6 GB) — and select a Linux image, Ubuntu or Debian, for your generation stack.
Install the OS yourself
Through the panel you deploy the chosen OS on a KVM server yourself — the environment stays fully under your control.
Set up the WebUI
Over SSH or the KVM console install Python and the WebUI, load your models and open the generation interface in the browser.
A server for your generation
We will suggest which plan fits your model library and generation speed: RAM, vCPU count and NVMe space for checkpoints. For fast GPU generation we will put together a configuration separately.
We will help set up the WebUI and secure remote access. Configurations with a graphics card are on the GPU servers page; an overview of ML workloads is on AI/ML VPS.
Help me size itFrequently asked questions
Didn't find your answer?
Write to us via the contact page.
Any — AUTOMATIC1111, ComfyUI and others. You have root access and install the interface, models and extensions yourself.
Yes. Load any checkpoints, LoRA and VAE — fast NVMe drives keep large weights right next to the generation.
The base plans focus on vCPU, RAM and NVMe. For fast generation with a graphics card we will put together a GPU configuration separately.
Yes. Once the WebUI is running it opens at the server address — generation happens on the VPS, not on your PC.
Technically yes, but on a CPU one image takes minutes instead of seconds. Practical work needs a GPU configuration — we build one per task.
One SDXL model is 6 GB or more, and with add-ons and LoRA a library quickly grows to hundreds of gigabytes. Take 400 GB NVMe or more (Max 1, €42.99/mo).
Only behind a password and TLS: an openly accessible generation interface gets found and abused quickly. Put Nginx with authentication in front of the app.
By default no: installation takes a couple of minutes and is done from the panel. If you need a turnkey setup, that is a server administration service from €79/mo.