48 GB · Inference · last run 2 Nov, 20:08 UTC
Cheapest by term: . On demand, the median is $0.91. Every chip's reserved prices →
Each provider's price for each commitment it states, against its own on-demand price. USD per GPU-hour.
| Provider | Term | Per GPU-hour | Off its on-demand |
|---|---|---|---|
| Hyperstack | not stated | $0.70 | −30% |
Each dot is one provider's own median. Hollow: shown, not counted.