Evidence labels reference
The rights, freshness, and confidence labels we apply to every row, defined and explained.
Providers
AI GPU Provider Directory
AI GPU providers we link to a public price page.
Compare AI GPU providers using rows that each link to a public price page on the provider's own site. Every visible row carries an editorial check on the source and a recent check timestamp. Rows whose source has not been editorially checked yet are kept in the data but not shown here.
Providers
Providers in the directory today
8 provider rows link to a public price page that has been editorially checked and is within the 30-day freshness cap. Additional rows appear here as more sources are reviewed.
| Provider | Segment | GPU | Mode | Region | Last checked | Source |
|---|---|---|---|---|---|---|
| AWS | Hyperscalers | H100 | On Demand | us-east-1 | May 28, 2026 | https://aws.amazon.com/ec2/pricing/on-demand/ |
| Azure | Hyperscalers | H100 | On Demand | eastus | May 28, 2026 | https://azure.microsoft.com/en-us/pricing/details/virtual-machines/linux/ |
| CoreWeave | AI-native clouds | H100 | On Demand | us | May 28, 2026 | https://docs.coreweave.com/pricing/pricing-instances |
| Crusoe | AI-native clouds | H100 | On Demand | Multi-region | May 28, 2026 | https://www.crusoe.ai/cloud/pricing |
| Hyperstack | Developer GPU clouds | H100 | On Demand | Multi-region | May 28, 2026 | https://www.hyperstack.cloud/gpu-pricing |
| Lambda | Developer GPU clouds | H100 | On Demand | Multi-region | May 28, 2026 | https://lambda.ai/pricing |
| Oracle | Hyperscalers | H200 | On Demand | us-ashburn-1 | May 28, 2026 | https://www.oracle.com/cloud/compute/gpu/pricing/ |
| RunPod Secure | Developer GPU clouds | H100 | On Demand | Multi-region | May 28, 2026 | https://www.runpod.io/pricing |
About the list
How a row gets here
Every row links a public price page on the provider site, has been editorially checked, and is within a 30-day recheck window. Rows whose source has not been editorially checked stay in the data but do not appear above. The full label vocabulary lives on the evidence labels reference.
Use the directory
Turn provider information into a plan
The directory tells you where capacity can be requested. The calculators turn that into a budget, and Compute College explains the trade-offs behind the numbers.
GPU-Hour Cost Calculator
Estimate AI compute cost from GPU price, runtime, utilization, and overhead.
How to compare GPU cloud quotes
Compare rate, reliability, network, and completed-workload cost.
On-demand vs reserved vs spot GPU pricing
Compare access terms, savings, and interruption risk.
Evidence labels reference
The rights, freshness, and confidence labels we apply to every row, defined and explained.