Provider pricing
On-demand and spot GPU cloud with hostnode-level pricing.
Prices below are live, pulled directly from the same feed Badgr uses to source overflow capacity.
| GPU | Lowest $/GPU-hr | Avg $/GPU-hr | Offers seen | Regions |
|---|---|---|---|---|
| H100 80GB | No live offers | — | — | — |
| A100 80GB | No live offers | — | — | — |
| L40S 48GB | No live offers | — | — | — |
| A6000 48GB | No live offers | — | — | — |
| RTX 4090 24GB | No live offers | — | — | — |
Data last refreshed within the last 5 minutes. Prices are third-party offers observed by Badgr, not Badgr's own rate card.
Badgr routes to whichever source — including TensorDock when it's the cheapest live option — has the best available price, with max-cost protection and one API regardless of who's underneath.
L4 24GB · United States
from $0.10/hr
L4 24GB · Europe
from $0.10/hr
RTX 5090 32GB · United States
from $0.25/hr