A cloud GPU machine in 4 steps

1Create an account and add credits

Sign up with your email and phone number. Add credits from $1.12. Credits never expire.

2Choose your GPU tier

Pick a tier by the GPU memory your work needs. Not sure? Start with Standard (RTX 3090, 24 GB VRAM).

3Open Studio or connect over SSH

Your machine is usually ready in a few minutes. Open Studio (JupyterLab) in your browser with one click, or use the SSH command on your machine page. PyTorch and CUDA are preinstalled.

4Stop when you’re done — the charges stop

Hit Stop in your dashboard. The machine’s disk is kept for 24 hours, then deleted.

Connection Methods

Studio (JupyterLab)

Best For
Notebooks, a terminal, uploading and downloading files
Setup
None — opens in your browser

SSH

Best For
Scripts, long-running jobs, copying files with scp
Setup
Add your public key once, from a Studio terminal

What You Need

  • ✓A computer with a current web browser — For Studio (JupyterLab)
  • ✓An SSH client (optional) — Built into macOS and Linux; OpenSSH or PuTTY elsewhere
  • ✓A GamePC account — With at least $1.00 credit to deploy
Get Started