Coming soon
Compare GPU clouds by the cost of a finished job.
Hourly prices are a starting point. The useful comparison includes storage, data transfer, idle time, stock and the work completed by the end of the run.
Price per hour
Comparable GPU, region and commitment rates.
Costs around the GPU
Storage, transfer, idle time, availability and setup.
Cost per finished job
Our own repeatable renders and inference workloads.
First comparison set
Runpod, Lambda, Vast.ai, Modal and others.
The first release will compare providers on the same workload and show the assumptions beside every result.
Demand test
Get notified when it's live
Your submit helps decide how soon the GPU comparison is built.