Community
Independent GPUs. A private VM each.
Priced by their owners. Your VM, your GPU.
- VM
- dedicated to this session
- GPU
- attached to this VM only
- Access
- SSH · JupyterLab
- At the end
- disk erased
Every session
Isolated by a VM. Not a container.
A VM of your own
Nothing else runs in it.
The GPU, yours alone
Never shared, never sliced.
Nothing left behind
Work disk erased when you stop.
A session
Catalog to notebook. Four steps.
- 01
Pick a machine
GPU, performance, country.
- 02
Start
Your VM boots with your key.
- 03
Connect
SSH or JupyterLab.
- 04
Stop
Pay by the minute. Nothing left behind.
The limit
Know what Community protects.
The provider can read a session's memory. Confidential data? Choose Secure.
- Good fit
- Open data, research, rendering, open models
- Choose Secure
- Confidential code, private or customer data
- Same account
- One login and one bill for both tiers
Own a GPU machine?
Rent it out on Community, at your price.
Request access.
Tell us what you will run. Your access arrives by email.