Now offering crypto payments →

GPU Cloud Pricing · Hopper

H100 PCIe Cloud Pricing

Hopper compute in a standard PCIe slot — typically the cheapest way to rent H100-class silicon when you don't need NVLink node scaling.

In stock — 8 GPUs available nowFrom $1.87/GPU/hr
Live pricing

H100 PCIe price per GPU-hour

GPU.ai rates are the live fleet floor — the cheapest provider with stock, passed through at cost. Competitor rates are public on-demand list prices.

Provider$ / GPU / hr
GPU.ai — Secure Cloud$2.48
GPU.ai — Community Cloud$1.87
CoreWeaveNot available
AWSNot available
AzureNot available

Need scale? Real 8× H100 PCIe nodes are in inventory from $2.00/GPU/hr.

Available regions: Asia Pacific · US Central · US East

How these prices are computed

GPU.ai prices are read live from our aggregated fleet of 12+ GPU clouds and normalized to a per-GPU hourly rate (a node's price is divided by its GPU count). We charge the provider's raw rate with zero markup and bill per second. Competitor figures are each provider's public on-demand list rate for the closest equivalent SKU — never spot, reserved, or committed-use discounts. Stock counts reflect launchable inventory, not theoretical capacity. Last updated 2026-08-17; this page refreshes hourly.

Specifications

NVIDIA H100 PCIe specs

ArchitectureHopper
VRAM80 GB HBM2e
Memory bandwidth2,000 GB/s
FP16 tensor (dense)756 TFLOPS
CUDA cores14,592
Power (TDP)350 W
InterconnectPCIe 5.0
Released2022
What fits

What runs on 80 GB?

Minimum H100 PCIe count needed to hold popular open models in VRAM, at FP16 and 4-bit quantization.

ModelParamsFP164-bit
Llama 3.1 8B8B1× H100 PCIe1× H100 PCIe
Mistral 7B7B1× H100 PCIe1× H100 PCIe
Gemma 2 27B27B1× H100 PCIe1× H100 PCIe
Qwen2.5 32B32B1× H100 PCIe1× H100 PCIe
Llama 3.1 70B70B4× H100 PCIe1× H100 PCIe
Qwen2.5 72B72B4× H100 PCIe1× H100 PCIe
Mixtral 8x22B141B MoE8× H100 PCIe2× H100 PCIe
Llama 3.1 405B405BDoesn't fit4× H100 PCIe
FLUX.1 dev12B1× H100 PCIe1× H100 PCIe
SDXL3.5B1× H100 PCIe1× H100 PCIe
Whisper large-v31.5B1× H100 PCIe1× H100 PCIe

Weights-only estimates with a 20% margin for KV cache and activations at moderate context. Long contexts, large batches, and training need substantially more memory.

FAQ

H100 PCIe rental questions

How much does it cost to rent an H100 PCIe in the cloud?

H100 PCIe cloud instances on GPU.ai currently start at $1.87 per GPU per hour, billed per second. That is the live floor price across every provider in our fleet — we route your launch to the cheapest provider with stock and add zero markup.

How much VRAM does the H100 PCIe have?

The NVIDIA H100 PCIe has 80 GB of HBM2e memory with up to 2,000 GB/s of bandwidth. GPU.ai reports the minimum VRAM actually delivered across our fleet, so the figure you see at launch is the figure you get.

Can I rent a multi-GPU H100 PCIe node?

Yes — real 8× H100 PCIe nodes are launchable right now, from $2.00 per GPU per hour. We only advertise node sizes that exist in live inventory, not theoretical configurations.

How does GPU.ai's H100 PCIe pricing compare to AWS, Azure, and CoreWeave?

GPU.ai aggregates a dozen-plus GPU clouds and passes through each provider's raw price with no markup, so the same silicon is typically far cheaper than hyperscaler list rates. The comparison table on this page shows the current public on-demand list prices side by side.

How does billing work?

Billing is per second, starting when your instance is SSH-ready and stopping the moment you terminate. There are no minimum commitments, no reservation fees, and no egress surprises — you pay the listed hourly rate pro-rated to the second.