Launch

Launch pricing: up to 74% below on-demand, fixed for your term — ends Oct 31, 2026 · 42d left Launch pricing ends Oct 31 · 42d left

See pricing
Deploy a GPUDeploy

Deploy

Deploy a GPU server

In stock · No KYC · Paid in crypto · Live in under 10 minutes once your payment confirms.

  1. GPU
  2. 2Configure
  3. 3Account
  4. 4Review & pay
Step 2 of 4 · ConfigureNext: Account

Configure your server

Sensible defaults are pre-selected — keep them all and continue, or tune what matters to you.

Billing term

Launch pricing ends Oct 31, 2026 — your rate stays fixed for the whole term.

Region

All sites are Tier III, ISO 27001-certified data centers with the same pricing.

Operating system imageSelected: Ubuntu 24.04 LTS

Every image ships with the NVIDIA driver, CUDA and Docker unless stated otherwise; ML images add JupyterLab on port 8888.

Linux

ML frameworks

Inference

Containers

Windows

Windows images are available on workstation-class GPUs only (RTX PRO 6000, RTX 5090, RTX 4090).

Custom

Extra block storage

4 TB of local NVMe is included. Network volumes cost $25 per TB per month and can be resized later.

Additional IPv4

One dedicated IPv4 and a /64 IPv6 block are included. Extra addresses cost $4 each per month.

Access

How you log in the minute the server is live.

Where do I find my public key?

macOS / Linux — in a terminal: cat ~/.ssh/id_ed25519.pub

Windows — in PowerShell: type $env:USERPROFILE\.ssh\id_ed25519.pub

No key yet? Run ssh-keygen -t ed25519, press Enter through the prompts, then run the command above.

Copy the whole line — it starts with ssh-ed25519 or ssh-rsa. Never paste a file that begins with -----BEGIN: that is your private key.

Server detailsoptional

Letters, digits and dashes. Left blank, we assign gpu-01.
Primary workload
Pick one and we pre-tune NCCL, drivers and the container runtime for it — click again to clear.
Back
Total per month$3,601