Exact public Linux on-demand EC2 prices in US East (N. Virginia), normalized by documented physical GPU count.
Compute
Azure GPU VM retail prices
$3.673 / GPU-hr
Azure NC24ads A100 v4 ranks first
Normalized retail price per GPU
6
GPU types ranked
Jul 27
Snapshot recorded
2026
All 6 GPU types by normalized retail price per GPU
Azure NC24ads A100 v4 is lowest at $3.673 / GPU-hr; the middle of the board sits at $6.98.
- Highest published row
- $12.290 / GPU-hr
- Middle of the board
- $6.98
- 1Azure NC24ads A100 v4$3.673 / GPU-hr
- 2Azure ND96amsr A100 v4$4.096 / GPU-hr
- 3Azure NC40ads H100 v5$6.980 / GPU-hr
- 4Azure NC80adis H100 v5$6.980 / GPU-hr
- 5Azure ND96isr H200 v5$10.600 / GPU-hr
- 6Azure ND96isr H100 v5$12.290 / GPU-hr
azure gpu: 8 on 3 of 6 rows · azure gpu memory per device gb: 80 on 3 of 6 rows · 5 checks behind each row.
| Rank | Name | Normalized retail price per GPU | Azure GPU memory bandwidth tb s | Azure GPU VM retail price usd hour | Why |
|---|---|---|---|---|---|
| 01 | Azure NC24ads A100 v4 | $3.673 / GPU-hr | 1.9 | $3.67 | Showing |
| 02 | Azure ND96amsr A100 v4 | $4.096 / GPU-hr | 2 | $32.77 | This rank |
| 03 | Azure NC40ads H100 v5 | $6.980 / GPU-hr | 3.9 | $6.98 | This rank |
| 04 | Azure NC80adis H100 v5 | $6.980 / GPU-hr | 3.9 | $13.96 | This rank |
| 05 | Azure ND96isr H200 v5 | $10.600 / GPU-hr | 4.8 | $84.8 | This rank |
| 06 | Azure ND96isr H100 v5 | $12.290 / GPU-hr | 3.4 | $98.32 | This rank |
Why this rank
Azure NC24ads A100 v4 on Azure GPU VM Retail Prices
Some explanation inputs are unavailable. Not published: comparable prior ranking.
Ranks a fixed Azure GPU VM cohort by comparable public Linux on-demand retail price per physical GPU.
1
Rank on this board
3.673 USD/GPU hour
Score
5
Inputs behind the score
100% confidence · 1 of 1 components evidenced
What the score is built from
1 of 1 evidencedEvery component evidenced · 1 evidence record each
| Component | Value | Contribution |
|---|---|---|
| Azure GPU normalized retail price usd hour | 3.673 | 3.673 |
New in this snapshot
No change in rankno_prior_published_snapshot
No component moved since the previous snapshot.
Where each number comes from
5 public recordsPhysical GPU countAzure GPU · observed Jul 27, 2026, 12:00 AM UTC1 GPUs100% confidence
- Source
- Microsoft Azure Retail Prices API
- Grade
- A
- Freshness
- fresh
- Locator
{"skuId":"DZH318Z09TGJ/001K","filter":"armRegionName eq 'westus3' and armSkuName eq 'Standard_NC24ads_A100_v4' and priceType eq 'Consumption'","meterId":"02e920ad-0b32-5221-84d0-99b6e54c215d","endpoint":"https://prices.azure.com/api/retail/prices","apiVersion":"2023-01-01-preview","specificationUrl":"https://learn.microsoft.com/en-us/azure/virtual-machines/sizes/gpu-accelerated/nca100v4-series","acceleratorSpecificationUrl":"https://www.nvidia.com/en-us/data-center/a100/"}
Memory bandwidth per GPUAzure GPU memory bandwidth tb s · observed Jul 27, 2026, 12:00 AM UTC1.935 TB/s100% confidence
- Source
- Microsoft Azure Retail Prices API
- Grade
- A
- Freshness
- fresh
- Locator
{"skuId":"DZH318Z09TGJ/001K","filter":"armRegionName eq 'westus3' and armSkuName eq 'Standard_NC24ads_A100_v4' and priceType eq 'Consumption'","meterId":"02e920ad-0b32-5221-84d0-99b6e54c215d","endpoint":"https://prices.azure.com/api/retail/prices","apiVersion":"2023-01-01-preview","specificationUrl":"https://learn.microsoft.com/en-us/azure/virtual-machines/sizes/gpu-accelerated/nca100v4-series","acceleratorSpecificationUrl":"https://www.nvidia.com/en-us/data-center/a100/"}
Memory per GPUAzure GPU memory per device gb · observed Jul 27, 2026, 12:00 AM UTC80 GB100% confidence
- Source
- Microsoft Azure Retail Prices API
- Grade
- A
- Freshness
- fresh
- Locator
{"skuId":"DZH318Z09TGJ/001K","filter":"armRegionName eq 'westus3' and armSkuName eq 'Standard_NC24ads_A100_v4' and priceType eq 'Consumption'","meterId":"02e920ad-0b32-5221-84d0-99b6e54c215d","endpoint":"https://prices.azure.com/api/retail/prices","apiVersion":"2023-01-01-preview","specificationUrl":"https://learn.microsoft.com/en-us/azure/virtual-machines/sizes/gpu-accelerated/nca100v4-series","acceleratorSpecificationUrl":"https://www.nvidia.com/en-us/data-center/a100/"}
Normalized retail price per GPUAzure GPU normalized retail price usd hour · observed Jul 27, 2026, 12:00 AM UTC3.673 USD/GPU hour100% confidence
- Source
- Microsoft Azure Retail Prices API
- Grade
- A
- Freshness
- fresh
- Locator
{"skuId":"DZH318Z09TGJ/001K","filter":"armRegionName eq 'westus3' and armSkuName eq 'Standard_NC24ads_A100_v4' and priceType eq 'Consumption'","meterId":"02e920ad-0b32-5221-84d0-99b6e54c215d","endpoint":"https://prices.azure.com/api/retail/prices","apiVersion":"2023-01-01-preview","specificationUrl":"https://learn.microsoft.com/en-us/azure/virtual-machines/sizes/gpu-accelerated/nca100v4-series","acceleratorSpecificationUrl":"https://www.nvidia.com/en-us/data-center/a100/"}
VM retail priceAzure GPU VM retail price usd hour · observed Jul 27, 2026, 12:00 AM UTC3.673 USD/hour100% confidence
- Source
- Microsoft Azure Retail Prices API
- Grade
- A
- Freshness
- fresh
- Locator
{"skuId":"DZH318Z09TGJ/001K","filter":"armRegionName eq 'westus3' and armSkuName eq 'Standard_NC24ads_A100_v4' and priceType eq 'Consumption'","meterId":"02e920ad-0b32-5221-84d0-99b6e54c215d","endpoint":"https://prices.azure.com/api/retail/prices","apiVersion":"2023-01-01-preview","specificationUrl":"https://learn.microsoft.com/en-us/azure/virtual-machines/sizes/gpu-accelerated/nca100v4-series","acceleratorSpecificationUrl":"https://www.nvidia.com/en-us/data-center/a100/"}
More in Compute
GPU rental prices, availability, installed base, and accelerator economics.
Official on-demand attached-GPU component prices in Iowa, with host VM CPU, memory, storage, and networking explicitly excluded.
Hourly Akash Network median and minimum GPU-hour prices, available units, total supply, provider count, memory, and interface.
See how often AI names you.
Enter your website. In about two minutes, Rank.ai asks ChatGPT, Claude and Gemini 12 questions your customers ask and grades how often they name you.
- Your grade out of 100How often AI names you, cites your site, and how high it ranks you.
- Who gets namedEvery competitor in the answers, most named first.
- The pages AI readsThe sources behind each answer.
- Three fixesWhat to fix first, with a brief for the first article.