How much does an NVIDIA H100 cost per hour in the cloud?

On-demand H100 rates range from about $2–$3 per GPU-hour on budget providers like Thunder Compute, Vast.ai, RunPod, and Lambda Labs, up to $7–$13 per GPU-hour on hyperscalers such as Azure, Google Cloud, and AWS.

Why are Thunder Compute’s H100 prices lower than AWS or Google Cloud?

Thunder Compute focuses on single-GPU, developer-centric instances and runs lean infrastructure without long-term reservation mark-ups, allowing prices that are roughly 4–8× cheaper than hyperscalers.

When should I choose an H100 instead of an A100?

Pick an H100 when your workload needs features like the Transformer Engine, higher memory bandwidth, or larger models with FP8 precision. For most prototyping and fine-tuning jobs, the lower-cost A100 (80 GB) delivers better runtime per dollar.

Do the listed prices include networking, storage, or egress fees?

No. The comparison shows pure per-GPU compute cost. Network, storage, and data-egress charges vary by provider and are billed separately.

How often is this H100 pricing data updated?

Thunder Compute reviews public price lists and updates the comparison every quarter to keep the numbers current.

Does Thunder Compute offer startup credits or free trials?

Yes. New users receive $20 in free credit, and additional startup credits are available on request.

Back

NVIDIA H100 Pricing (May 2025): Cheapest On-Demand Cloud GPU Rates

Comparison of single-GPU H100 hourly costs on AWS, Azure, GCP, Lambda Labs, RunPod, Vast.ai, and other U.S. providers.

Published:

May 19, 2025

Last updated:

May 21, 2025

The table below compares current on-demand, hourly rental prices for a single NVIDIA H100 80GB GPU across major U.S. cloud providers. Prices are normalized per-GPU (even if only multi-GPU instances are offered) and reflect standard on-demand rates in U.S. regions (no spot, reserved, or non-US pricing).

Provider	SKU/Instance	On-Demand $/GPU-hr*	Notes
AWS	p5.48xlarge (8 × H100 80GB)	~$12.29	8-GPU instance (p5.48xl) at ~$98.32/hr total.
Azure	NC H100 v5 VM (1 × H100 80GB)	$6.98	Single H100 GPU VM in East US region.
Google Cloud	A3 High (a3-highgpu-1g, 1 × H100)	~$11.06	1× H100 80GB in us-central (on-demand).
Oracle Cloud	BM.GPU.H100.8 (8 × H100 80GB)	~$2.98	8-GPU bare-metal node; ~$2.98 per GPU on-demand.
Lambda Labs	8× NVIDIA H100 SXM (80GB)	$2.99	8-GPU Lambda Cloud instance on HGX system.
CoreWeave	8× H100 HGX (80GB, InfiniBand)	~$6.16	8-GPU HPC node w/ InfiniBand, $49.24/hr total.
Paperspace	H100 80GB (Dedicated instance)	$5.95	On-demand price per H100; multi-GPU discounts available.
RunPod	H100 80GB (PCIe)	$1.99	Community cloud price (per GPU); Secure Cloud is $2.39/hr.
Vast.ai	H100 80GB (various hosts)	~$2.07	Marketplace lowest current price per H100 GPU.

*Normalized cost per single H100 GPU, even when only multi-GPU instances are offered by the provider.

Methodology (why you can trust these numbers)

On-demand only: No reserved-instance, commitment, or prepaid discounts.
Same class of silicon: All prices refer to NVIDIA H100 80GB GPUs. Thunder Compute’s A100 80GB rate is also shown to help developers evaluate cost-performance tradeoffs.
Public price lists: Every figure comes from the provider’s current pricing page (or public documentation) on the date above; where a provider sells only 8-GPU nodes we divide by eight to get a single-GPU equivalent.
USD in U.S. regions: Rates elsewhere can differ by 5-20 %.

A100 cost-performance benchmark

Thunder Compute (A100)

A100 80GB**

$0.78

Not an H100; shown for cost-performance comparison

**Thunder Compute’s A100 80GB rate is included for reference. While A100s are a generation older, they remain highly capable for many workloads and offer dramatically better cost-efficiency for prototyping, fine-tuning, and small-scale training.

Why this matters for developers

Provider	2 hrs runtime	Effective cost
Thunder Compute – A100 80 GB	2 × $0.78	$1.56
Vast.ai	2 × $2.07	$4.14
RunPod	2 × $2.69	$5.38
Lambda Labs	2 × $2.99	$5.98
Azure	2 × $6.98	$13.96
Google Cloud	2 × $11.06	$22.12
AWS	2 × $12.29	$24.58

Price sources: Thunder Compute pricing page, Lambda Labs “GPU Cloud” grid, RunPod pricing, Vast.ai median market price, and the DataCrunch hyperscaler comparison for AWS, Google Cloud, and Azure. vast.ai

Result: Two hours on Thunder Compute’s A100 costs less than 15 minutes on AWS or GCP H100s—and the A100 still gives you roughly 15× more runtime per dollar than hyperscaler H100s.

Takeaways

Thunder Compute’s A100 rate is 4–8× cheaper than AWS or GCP and ≈2× cheaper than Azure, while its A100 remains the absolute price-performance leader.
Specialized providers like Vast.ai, RunPod, and Lambda have narrowed the gap, but they still charge 2–3× more than Thunder Compute for equivalent runtime.
Unless your workload truly needs H100 features (Transformer Engine, higher bandwidth, etc.), the A100 often delivers the best ROI for prototyping, fine-tuning, and small-scale training.
Bookmark this page—we refresh the numbers quarterly so you don’t have to.
Building a startup? See our analysis of Startup-Friendly GPU Cloud Providers for credit offers, and spin up an A100 or H100 on Thunder Compute

Carl Peterson

Try Thunder Compute

Start building AI/ML with the world's cheapest GPUs

NVIDIA H100 Pricing (May 2025): Cheapest On-Demand Cloud GPU Rates

Methodology (why you can trust these numbers)

A100 cost-performance benchmark

Why this matters for developers

Takeaways

Other articles you might like