What renters pay for an H100

H100 rental price per GPU-hour, from the providers' own pages

Updated

All but 2 of the 8 providers we read publish an on-demand H100 price per GPU-hour.

Describe my GPUs

On demand, by provider

Each provider's H100 price is in the H100 column of the home page's price table, and the sentence it was read from is on the board on the dataset page, What GPU clouds charge per GPU-hour. The spread is wide: marketplaces built on bids sit beside data-centre clouds that price a whole eight-card machine.

Reserved and spot

6 providers state reserved, spot or commitment terms, from committed-spend pricing below list to no commitment at all.

Tell us about the GPUs you could rent out

Tell us the cards, how many and where they are. We reply with what renters pay for them on the table and which marketplaces list third-party hosts.

Your GPUs

Which GPUs do you have?

  • Every price quoted from a provider's own page
  • We never take or resell capacity

Sources

This site's own datasets, every figure quoted from its source and dated:

  • The GPU Host Payouts by Card carries 7 published figures read from 2 sources' own pages; the 6 published figures given by card run from 0.14 to 1.44 USD, read 22 October 2026.
  • What GPU clouds charge per GPU-hour: 7 of 8 GPU rental providers publish a per-GPU-hour price on the pages we read, read 8 October 2026.
  1. Vast.ai pricing
  2. RunPod pricing
  3. RunPod GPU instance pricing
  4. Lambda pricing
  5. CoreWeave pricing
  6. Nebius prices
  7. Together AI pricing
  8. SaladCloud pricing

Got GPUs sitting idle?

Tell us the cards, how many and where they are. We reply with what renters pay for them on the table and which marketplaces list third-party hosts.

Describe my GPUs