H200 rental prices per GPU-hour
Hopper with 141 GB HBM3e. Mostly SXM / HGX. Collected 2026-11-10 from the vendors' own pages. 23 price points, 14 providers.
| Provider | Secure | Community | Market | Serverless | Spot | Reserved |
|---|---|---|---|---|---|---|
| RunPod | $5.29 | $3.59 | · | · | · | · |
| Vast.ai | · | · | from $3.09 | · | · | · |
| Nebius | $5.40 | · | · | · | from $0.79 | · |
| Hyperstack | $3.99 | · | · | · | · | from $2.79 |
| Verda (ex-DataCrunch) | $6.44 | · | · | · | $3.22 | · |
| Jarvis Labs | $4.59 | · | · | · | $2.29 | · |
| Together AI | $5.99 | · | · | · | $2.99 | · |
| DigitalOcean GPU Droplets | $4.47 | · | · | · | · | from $3.40 |
| CoreWeave | $6.31 | · | · | · | $2.62 | · |
| Modal | · | · | · | $4.54 | · | · |
| Crusoe | $4.29 | · | · | · | · | · |
| AWS | $7.91 | · | · | · | · | · |
| Microsoft Azure | $10.60 | · | · | · | · | · |
cooler = cheaper warmer = pricier (scaled within this map, reserved excluded). Purple dashed = "from" price, committed term or announced change. Not every provider has every tier or form.
All H200 offers
Rows ordered by price within each tier. The form column is mandatory: compare only matching forms.
| Tier | Provider | Form | $ / GPU-hr | Notes | Source |
|---|---|---|---|---|---|
| Secure | Hyperstack | SXM (HGX-class) | $3.99 | price page | |
| Secure | Crusoe | SXM (HGX-class) | $4.29 | HGX H200; new price $5.40 from Dec 1, 2026 | price page |
| Secure | DigitalOcean GPU Droplets | SXM (HGX-class) | $4.47 | price page | |
| Secure | Jarvis Labs | SXM (HGX-class) | $4.59 | price page | |
| Secure | RunPod | SXM (HGX-class) | $5.29 | price page | |
| Secure | Nebius | SXM (HGX-class) | $5.40 | price page | |
| Secure | Together AI | SXM (HGX-class) | $5.99 | GPU Clusters, on-demand | price page |
| Secure | CoreWeave | SXM (HGX-class) | $6.31 | per GPU-hour from the 8-GPU node price ($50.44) | price page |
| Secure | Verda (ex-DataCrunch) | SXM (HGX-class) | $6.44 | price page | |
| Secure | AWS | SXM (HGX-class) | $7.91 | p5en.48xlarge/8, us-east-1 | price page |
| Secure | Microsoft Azure | SXM (HGX-class) | $10.60 | ND96isr H200 v5/8, East US 2 | price page |
| Community | RunPod | SXM (HGX-class) | $3.59 | price page | |
| Market | Vast.ai | SXM (HGX-class) | from $3.09 median $4.34 | price page | |
| Market | Vast.ai | NVL / NVLink-bridged | from $3.33 median $3.99 | price page | |
| Serverless | Modal | SXM (HGX-class) | $4.54 | $0.001261 per second × 3,600 | price page |
| Serverless | Replicate | form not stated | $5.49 | gpu-h200, $0.001525 per second; committed-spend capacity only | price page |
| Spot | Nebius | SXM (HGX-class) | from $0.79 | preemptible, "from" | price page |
| Spot | Jarvis Labs | SXM (HGX-class) | $2.29 | price page | |
| Spot | CoreWeave | SXM (HGX-class) | $2.62 | spot per GPU-hour from the 8-GPU node price ($20.93) | price page |
| Spot | Together AI | SXM (HGX-class) | $2.99 | preemptible compute | price page |
| Spot | Verda (ex-DataCrunch) | SXM (HGX-class) | $3.22 | price page | |
| Reserved | Hyperstack | SXM (HGX-class) | from $2.79 | reservation "starting from" | price page |
| Reserved | DigitalOcean GPU Droplets | SXM (HGX-class) | from $3.40 | 12-month reserved price | price page |
Questions
How much does it cost to rent an H200?
On 2026-11-10: secure on-demand SXM-class prices ranged $3.99 to $10.60 per GPU-hour across 11 providers; community and marketplace prices were lower and spot lower still. Form and tier matter, see the table.
What is the difference between SXM and PCIe H200?
SXM cards sit on HGX boards with a fast GPU-to-GPU fabric and are usually priced higher; PCIe cards are standard add-in cards. We never rank them against each other without a label.
Is spot cheaper than on-demand?
Per hour, yes, but spot can be interrupted. The estimator adds your own rerun rate so you can compare the whole job.
Sponsored row: none yet. Advertise on this page (waitlist). Sponsored rows would sit above the table and never change the order. Total a job.