Question 1

What is Thunder Compute?

Accepted Answer

Thunder Compute is a budget-friendly, on-demand GPU cloud platform built for AI/ML prototyping. It lets startups, researchers, and indie developers spin up powerful GPUs in seconds and connect with one click via a VS Code or Cursor extension.

Question 2

Which GPU instance types can I launch on Thunder Compute?

Accepted Answer

Current on-demand options include NVIDIA Tesla T4, A100 40 GB, A100 80 GB, and H100 GPUs, all provisioned in U.S. Central data centers.

Question 3

How much does an NVIDIA A100 40 GB instance cost per hour?

Accepted Answer

Just $0.66 per GPU-hour with no commitments or upfront reservations.

Question 4

How much does an NVIDIA A100 80 GB instance cost per hour?

Accepted Answer

$0.78 per GPU-hour on a pure pay-as-you-go basis.

Question 5

How much does a Tesla T4 instance cost per hour?

Accepted Answer

$0.27 per GPU-hour with the same on-demand flexibility.

Question 6

Do I need a long-term contract to get these prices?

Accepted Answer

No. All pricing is on-demand—launch and stop instances whenever you like and pay only for the minutes you use.

Blogs

Get the world's
cheapest GPUs

Blogs

Get the world's cheapest GPUs

Get the world's
cheapest GPUs