Is the premium for the NVIDIA H200 worth it? With its 141GB of HBM3e memory. This GPU is designed for high-throughput inference and memory-intensive training, but that performance comes with a significant price tag.
In this guide, we break down the NVIDIA H200 pricing landscape for February 2026. We compare on-demand rates across hyperscalers like AWS and Azure alongside specialized GPU clouds like Lambda and RunPod to help you find the most cost-effective path for your workloads.
Key Takeaways
<ul><li><strong>H200 premiums remain steep.</strong> Even after AWS's June price cut, the cheapest hyperscaler H200 hour costs 3 times more than Thunder Compute's H100.</li><li><strong>Specialist clouds narrow the gap.</strong> Lambda, RunPod, Jarvislabs, and Vast.ai all sit in the $2–4 range, but Thunder's H100 is still cheaper.</li><li><strong>Choose H200 only when you must:</strong> Running massive models that overflow 80 GB VRAM or long-context inference. For prototyping, fine-tuning, and most training, H100 80 GB wins on ROI.</li><li><strong>Thunder Compute roadmap.</strong> We don't offer H200 nodes yet; today you can launch H100 80 GB at $1.38/hr (one-click VS Code, per-second billing, persistent volumes, live hardware swaps).</li></ul>
One-minute snapshot
*Prices are normalized per single H200 even when a provider sells only 8-GPU nodes. U-S region pricing, on-demand only (no spot, reserved, or contract rates).
Methodology – why you can trust these numbers
<ul><li><strong>On-demand only.</strong> We excluded capacity reservations longer than 14 days, reserved instances, and spot/pre-emptible offers.</li><li><strong>Same silicon.</strong> Every row is a 141 GB NVIDIA H200 (SXM or PCIe).</li><li><strong>Public price lists only.</strong> Figures come straight from each provider's pricing page in February 2026.</li><li><strong>US regions, USD.</strong> Regional variation can add 5-20 percent; those are ignored for apples-to-apples comparison.</li></ul>
H100 vs H200 cost benchmark across generations
Bottom line: two hours on Thunder Compute's A100 costs less than 15 minutes of an H200 on Coreweave, and still buys roughly **13 × more runtime per dollar than hyperscaler H200s.