​
Select an offer

Start in the Compute Marketplace (sidebar). Choose a GPU type, apply any filters, then click Deploy on the offer you want. You’ll be taken to the deploy configuration page with the selected offer’s specs displayed at the top:
  • GPU model, VRAM, and architecture
  • Location
  • vCPU count, RAM, and included disk

​
Configure your deployment

1

Choose a template

Templates determine what is running on your instance when it boots:
  • Bare Metal — The plain VM: Ubuntu 22.04 with NVIDIA drivers and Docker, nothing else. SSH in and set up your own stack.
  • PyTorch + Jupyter — PyTorch with CUDA 12 and JupyterLab, opened in the browser.
  • Open WebUI + Ollama — Chat with open-source LLMs in the browser; pull any model from within the app.
  • ComfyUI — Node-based Stable Diffusion workflow editor for image generation.
See Templates for details on each.
2

Check the disk

Every QuantaCloud offer includes a fixed amount of local disk (for example, 256 GB on a single RTX A6000 and more on multi-GPU instances). It is shown with the offer specs and is included in the hourly rate, so there is nothing to configure. The disk is deleted when you stop the instance; there are no volumes or snapshots.
3

Choose an SSH key

Select which SSH key to use for the instance. Your managed (platform-generated) key is pre-selected by default.Keys are listed with their fingerprint for identification. If you need to add a new key, there’s a link to the SSH keys management page.
4

Review cost and deploy

The cost summary shows the offer’s hourly rate, with the disk included. Your current credit balance and estimated runtime (balance / hourly rate) are displayed.Click Deploy to launch. If your balance is insufficient, you’ll see Fund & Deploy, which takes you to add credits. When you’re back on the deploy page, check your template and click Deploy.

​
After deployment

Once you click Deploy, your instance goes through these states:
  1. Initializing — Setting up the deployment
  2. Provisioning — Creating your instance
  3. Connecting — Waiting for SSH access (and, for app templates, for the app to come up)
  4. Running — Your instance is ready to use
Most Bare Metal instances with 1 to 4 GPUs are running in about 3 minutes; 8-GPU instances take around 10 minutes. App templates take a few minutes longer while the container image is pulled. You can monitor progress in the Deployments page.
You’re charged for the first hour when you deploy, and for each further hour when the previous one is used up. If your balance can’t cover the next hour, the deployment is terminated and its disk is deleted. See Billing.

​
Stopping a deployment

In the Deployments page, click Stop on any running deployment. The instance is terminated, its disk is deleted, and the unused seconds of the current hour are refunded to your balance. If you stop it before it is running, the first hour is refunded in full. There is no way to pause an instance and keep its disk, so download anything you need first.

​
What happens if a deployment fails?

If provisioning fails, you’ll see a Failed status in the Deployments page with a short reason. The first hour is automatically refunded to your credit balance in full. You can try deploying again — consider picking a different offer or region if the issue persists.