from 16 GBImages
Images · Flux
Images from a text prompt: illustrations, covers, social posts and product shots.
from €0.16/h · RTX 4060 Ti, 16 GB
Run AI models on a rented GPU
Pick a task: images, video, chat or speech. The server installs the app and downloads the model, and the interface opens in your browser. Nothing to install on your computer, and you pay by the minute.
13 appsimages, video, chat and speech
2–5 minutesuntil the server is up
from €0.12/hcheapest suitable card
€0.03 per GBa month to keep your model folder
Ready-made apps: pick a task
A ready-made app is a server that comes with ComfyUI or a chat window and the right model already installed. The apps themselves are free; you only pay for GPU time. The price on each card is for the cheapest card with enough memory for that app.
Images and photos
Generate from a prompt or edit existing photos.
from 24 GBImages
Images · Qwen-Image
Posters, signs and banners with the lettering drawn right into the image.
from €0.30/h · RTX 3090, 24 GB
from 24 GBImages
Photo editor · Qwen Edit
Edit a photo with words: swap the background, turn day into evening, remove an object.
from €0.30/h · RTX 3090, 24 GB
from 12 GBImages
Images · SDXL
Thousands of community styles and LoRAs. Runs even on affordable 12 GB cards.
from €0.12/h · RTX 3060, 12 GB
from 24 GBImages
Images · HiDream-O1
Images from a prompt with a fresh open model. The MIT licence allows commercial use.
from €0.30/h · RTX 3090, 24 GB
from 48 GBVideo
Video · Wan 2.2 14B
High-quality clips from text or from a photo.
from €0.46/h · L40, 48 GB
from 24 GBVideo
Video · Wan 2.2 5B
A quick draft of a clip on an affordable card, to test the idea before the final version.
from €0.30/h · RTX 3090, 24 GB
Private ChatGPT
A chat window in your browser on Qwen3 and MiMo, no subscription.
from 24 GBChat
Private ChatGPT · Qwen3 32B
A chat in your browser for writing, code and documents. No subscriptions and no message limits.
from €0.30/h · RTX 3090, 24 GB
from 12 GBChat
Private ChatGPT · Qwen3 8B
A fast chat on an affordable card: email drafts, translations, quick answers.
from €0.12/h · RTX 3060, 12 GB
from 12 GBChat
Private ChatGPT · MiMo 9B
A fast chat from Xiaomi on an affordable card. It understands images too: send a photo or a screenshot.
from €0.12/h · RTX 3060, 12 GB
from 12 GBText to speech
Text to speech · Qwen3-TTS
Turns text into speech in English and other languages, with several ready-made voices.
from €0.12/h · RTX 3060, 12 GB
from 12 GBText to speech
Text to speech · VoxCPM2
Speech in 30 languages with natural intonation. Describe a voice in words or clone one from a sample.
from €0.12/h · RTX 3060, 12 GB
from 12 GBText to speech
Text to speech · Chatterbox
Speech in your own voice from a 5-second recording. 23 languages.
from €0.12/h · RTX 3060, 12 GB
How much video memory each app needs
The model has to fit entirely in the graphics card’s memory, so you choose by memory size. The lightest apps need 12 GB; the full Wan 2.2 14B video model needs 48 GB.
| Memory | Apps | Cheapest card | Per hour |
|---|---|---|---|
| from 12 GB | Images · SDXL, Private ChatGPT · Qwen3 8B, Private ChatGPT · MiMo 9B, Text to speech · Qwen3-TTS, Text to speech · VoxCPM2, Text to speech · Chatterbox | RTX 3060 12 GB | from €0.12 |
| from 16 GB | Images · Flux | RTX 4060 Ti 16 GB | from €0.16 |
| from 24 GB | Images · Qwen-Image, Photo editor · Qwen Edit, Images · HiDream-O1, Video · Wan 2.2 5B, Private ChatGPT · Qwen3 32B | RTX 3090 24 GB | from €0.30 |
| from 48 GB | Video · Wan 2.2 14B | L40 48 GB | from €0.46 |
For language models, memory depends on the number of parameters: 7–8B in 4-bit take about 5–6 GB, 70B about 50 GB, plus headroom for context. Every card and price is on the pricing page.
Run an AI model in 3 steps
- Sign up and top up. Create an account and top up from €10. To launch, your balance must cover at least one hour of the chosen machine.
- Choose the app. Click Launch on a card. The dashboard opens with the app and a suitable GPU already selected, and you can see the price and the machine’s country.
- Work in your browser. The server starts in 2–5 minutes, and the app downloads its model in a few more. Then a button in the dashboard opens ComfyUI, the chat or the speech app.
Done? Stop the server in the dashboard and billing stops. You can save your models and results with the /workspace folder; it will be on your next server automatically.
Your own model or code: setup guides
If the model you need is not among the ready-made apps, take an empty server: PyTorch and Jupyter, an empty ComfyUI, vLLM or your own Docker image. Each guide works out how much memory the model needs, which card fits and which commands to run.
Language model
DeepSeek
from 5 GB · from €0.12/h on RTX 3060
Language model
Llama
from 2 GB · from €0.12/h on RTX 3060
Language model
Qwen
from 5 GB · from €0.12/h on RTX 3060
Images and video
Stable Diffusion
from 6 GB · from €0.12/h on RTX 3060
Images and video
Flux
from 16 GB · from €0.16/h on RTX 4060 Ti
Tool
ComfyUI
from 6 GB · from €0.12/h on RTX 3060
Tool
Ollama
from 6 GB · from €0.12/h on RTX 3060
Tool
vLLM
from 6 GB · from €0.12/h on RTX 3060
Audio
Whisper
from 6 GB · from €0.12/h on RTX 3060
Images and video
Video generation
from 24 GB · from €0.30/h on RTX 3090
Questions about running AI models
What is a ready-made app?
A server that installs ComfyUI, the Open WebUI chat window or a text-to-speech app together with the right model when it starts: Flux, SDXL, Qwen-Image, HiDream-O1, Qwen Edit, Wan 2.2, Qwen3, MiMo, Qwen3-TTS, VoxCPM2 or Chatterbox. You open the interface in your browser and get to work. The apps are free; you only pay for GPU time.
Which GPU does an AI model need?
It depends on the model. SDXL, the Qwen3 8B and MiMo 9B chats and text to speech need 12 GB of VRAM, Flux needs 16 GB, Qwen-Image, HiDream-O1, Qwen Edit, Wan 2.2 5B and the Qwen3 32B chat need 24 GB, and Wan 2.2 14B video needs 48 GB. Table with prices
How long does a launch take?
The server usually starts in 2–5 minutes, then the app spends a few more minutes downloading the model. When everything is ready, a button in the dashboard opens the interface.
Do I need to install anything on my computer?
No. ComfyUI and the chat open in an ordinary browser, and all the work runs on the rented GPU. For your own code you have SSH and Jupyter.
What happens to my models and images after I stop?
When you stop, click Save and stop: the /workspace folder with your models, LoRAs and results goes to storage for €0.03 per GB per month and comes back on your next server automatically. If you stop without saving, the disk is deleted.
Can I use the results commercially?
It depends on the model’s licence. Flux.1 schnell, Qwen-Image, Wan 2.2, Qwen3-TTS and VoxCPM2 are under Apache 2.0, and HiDream-O1, MiMo and Chatterbox under MIT; both allow commercial use. Check the terms of any other model before you publish.
A server is usually ready in 2–5 minutes
Sign up in a minute, top up from €10, pay by the minute. Stop the server and billing stops.