NVIDIA H100 PCIe pricing.
prices as of · 2 providers · 3 configurations
The same chip as the SXM board on a slot card, with HBM2e instead of HBM3 and about 60% of its memory bandwidth.
The cheapest H100 PCIe today is H100 PCIe · Community Cloud at RunPod, $1.99/GPU-hr, read 17 Sept 2026; nobody publishes an interruptible rate for it.
Cheapest on demand
$1.99
per GPU-hour at RunPod
Median of the spread
$2.89
across 3 configurations
Cheapest interruptible
none
nobody publishes one
Per GB of memory
$0.0249/GB-hr
market median $0.0300/GB-hr
What the board is
published by NVIDIA, not by a provider- Memory
- 80 GB
- Memory bandwidth
- 2.00 TB/s
- Architecture
- Hopper
- Launched
- 2022
These are the chip's own figures, kept in this site's registry and the same whoever rents it out. A provider's feed is trusted for its prices, its GPU counts and the memory it states per card, never for what the board is. There is no speed score on this page: bandwidth is a published number, and turning it into a benchmark would be inventing one.
Every provider renting the H100 PCIe
Cheapest per GPU-hour first. The rate is for one GPU for one hour, so a node of eight sits beside a single card at a comparable number, with the whole machine's hourly price under it. Where a provider publishes an interruptible rate it is in the same cell.
RunPodH100 PCIe · Community Cloud
1x H100 PCIe · 80 GB each · 16 vCPU · 188 GB RAM
disk none, billed separately · regions not stated · billed per second · marketplace
Community Cloud: vetted third-party hosts, cheaper and less consistent · storage billed separately · billed per second
$1.99per GPU-hour$1.99/hr all inno spot rate
- RunPod
H100 PCIe · Secure Cloud
1x H100 PCIe · 80 GB each · 16 vCPU · 188 GB RAM
disk none, billed separately · regions not stated · billed per second
Secure Cloud: RunPod's own datacentres · storage billed separately · billed per second
$2.89per GPU-hour$2.89/hr all inno spot rate
Lambda1x NVIDIA H100 PCIe
1x H100 PCIe · 80 GB each · 26 vCPU · 225 GB RAM
disk 1 TB · regions not stated · billed per minute
Priced per GPU-hour, billed by the minute · storage included in the instance
$3.29per GPU-hour$3.29/hr all inno spot rate
Frequently asked
- Who rents the H100 PCIe cheapest?
- RunPod, at $1.99/GPU-hr on H100 PCIe · Community Cloud, read on 17 Sept 2026. That is the on-demand rate for one GPU for one hour; the whole machine bills $1.99/hr.
- What is the NVIDIA H100 PCIe?
- The same chip as the SXM board on a slot card, with HBM2e instead of HBM3 and about 60% of its memory bandwidth. It is a Hopper board from 2022 with 80 GB of memory at 2.00 TB/s, in the PCIe card package. Those figures come from the vendor, not from any provider's price page.
- Can I rent a H100 PCIe on spot?
- None of the 3 configurations here publishes an interruptible rate today. Where a provider does publish one, it sits beside the on-demand price in the table above.
- Why do H100 PCIe prices differ so much?
- The spread here runs from $1.99/GPU-hr to $3.29/GPU-hr for the same silicon, because what else comes with it differs: the vCPU and RAM around the GPU, whether a disk is included, how many regions it runs in, and whether the machine belongs to the provider or to a third party on a marketplace. Each row says which.
- How current are these prices?
- They are read every morning from each provider's own public page or API, and this page recomputes with them. Last refresh: 17 Sept 2026. Reserved and committed rates are not indexed. Prices exclude tax.