GPU price
NVIDIA B200 rental prices
The lowest listed on-demand price for an B200 is $6.00 per GPU-hour at Hyperstack, among 14 fresh listings from 7 providers. As of Oct 15, 2026.
Blackwell-generation GPU, sold mainly as 8-GPU HGX boards and DGX systems. Prices were last checked Oct 15, 2026. A listing not re-checked within 7 days is marked stale and left out of every lowest-price figure.
Lowest listed price by pricing type
| Pricing | $ per GPU-hour | Provider | Checked |
|---|---|---|---|
| On-demand | $6.00 | Hyperstack | Oct 15, 2026 |
| Spot | $0.99 | Nebius | Oct 15, 2026 |
| Reserved | $6.79 | Together AI, 91-180 days | Oct 15, 2026 |
On-demand price by provider
Each provider’s lowest fresh on-demand listing, $ per GPU-hour.
Every listed B200 price
| Provider | Configuration | Pricing | $ per GPU-hour | Region | Checked | Source |
|---|---|---|---|---|---|---|
| Nebius | Configuration not stated HGX B200 |
Spot | $0.99 listed: from $0.99 (preemptible) |
not stated | nebius.com | |
| Together AI | Configuration not stated HGX B200 |
Spot | $4.09 listed: $4.09 (listed as preemptible) |
not stated | together.ai | |
| CoreWeave | 8 GPUs per instance · 180 GB each HGX B200 |
Spot | $4.2638 from Spot Price: $34.11 / Hour |
North America | coreweave.com | |
| Hyperstack | 192 GB each B200 |
On-demand | $6.00 | not stated | hyperstack.cloud | |
| Lambda | SXM6 · 8 GPUs per instance · 180 GB each 8x B200 |
On-demand | $6.69 | not stated | lambda.ai | |
| Together AI | Configuration not stated HGX B200 |
Reserved, 91-180 days | $6.79 listed: $6.79 (reserved, 91-180 days) |
not stated | together.ai | |
| Lambda | SXM6 · 1 GPU per instance · 180 GB each 1x B200 |
On-demand | $6.99 | not stated | lambda.ai | |
| Together AI | Configuration not stated HGX B200 |
Reserved, 31-90 days | $7.79 listed: $7.79 (reserved, 31-90 days) |
not stated | together.ai | |
| Together AI | Configuration not stated HGX B200 |
Reserved, 3-30 days | $7.99 listed: $7.99 (reserved, 3-30 days) |
not stated | together.ai | |
| Together AI | Configuration not stated HGX B200 |
On-demand | $8.19 | not stated | together.ai | |
| Nebius | Configuration not stated HGX B200 |
On-demand | $8.50 | not stated | nebius.com | |
| CoreWeave | 8 GPUs per instance · 180 GB each HGX B200 |
On-demand | $8.60 from On-Demand Price: $68.80 / Hour |
North America and Europe | coreweave.com | |
| Oracle Cloud | 8 GPUs per instance · 180 GB each BM.GPU.B200.8 |
On-demand | $14.00 listed: $14 per GPU per hour (Compute - GPU - B200) |
not stated | apexapps.oracle.com | |
| AWS | 8 GPUs per instance p6-b200.48xlarge |
On-demand | $14.2416 from $113.9328 per instance-hour |
us-east-1 | aws.amazon.com |
Download: current prices (CSV) · current prices (JSON) · daily history (CSV) · price-change feed (RSS). Free to reuse with attribution under CC BY 4.0.
Price history
The history chart appears once there are two daily snapshots. Tracking began Oct 15, 2026; the first snapshot has the lowest listed on-demand price at $6.00 per GPU-hour.
Embed this chart
<iframe src="https://datacentercensus.com/embed/gpu-prices/b200" title="NVIDIA B200 rental prices" width="100%" height="500" style="border:0;max-width:640px" loading="lazy"></iframe>
<p style="font:13px/1.4 system-ui,sans-serif;margin:6px 0 0">Source: <a href="https://datacentercensus.com/gpu-prices/b200">Data Center Census</a></p>
<script>addEventListener("message",function(e){if(e.origin==="https://datacentercensus.com"&&e.data&&e.data.dcc){document.querySelectorAll('iframe[src^="https://datacentercensus.com/embed/"]').forEach(function(f){if(f.contentWindow===e.source)f.style.height=e.data.h+"px"})}})</script>Free with the credit line. Preview the embed.
NVIDIA B200 specifications
| Specification | Value |
|---|---|
| Architecture | Blackwell |
| NVLink, GPU to GPU | 1.8 TB/s per GPU |
| HGX B200 board (8 GPUs): total memory | 1.4 TB |
| HGX B200 board (8 GPUs): total NVLink | 14.4 TB/s |
| HGX B200 board (8 GPUs): FP8 tensor (sparse) | 72 PFLOPS |
| HGX B200 board (8 GPUs): FP16 / BF16 tensor (sparse) | 36 PFLOPS |
Source: NVIDIA product page. NVIDIA’s HGX page gives most B200 figures for the whole 8-GPU board, not per GPU, and the Census does not divide them. Throughput is stated with sparsity; NVIDIA says dense is half.
How these prices are collected
- What counts
- Only prices a provider publishes on a public pricing page or in its public price list, opened by the Census on the day shown. Marketplace offers, quotes and “contact sales” rows are not listed. Purchase prices are not listed at all.
- $ per GPU-hour
- The listed price per GPU-hour. Where a provider lists a price per instance, the Census divides it by the GPUs in that instance and shows the original figure under it. Nothing else is converted or estimated.
- Pricing types
- On-demand, spot (including tiers a provider calls preemptible or low-priority) and reserved with the term the provider states. Reserved rows without a stated term are not listed.
- Variants
- SXM, PCIe, NVL and similar versions are kept separate when the provider names them. A provider that does not name the variant is shown without one.
- Freshness
- Every listing carries a check date. 7 days after that date it is marked stale: still shown, but left out of “lowest” figures and the calculator’s default results.
- History
- Each daily run saves a snapshot, so the model pages can chart the lowest listed on-demand price over time and the change log can show what moved. History began Oct 15, 2026.
- Differences between providers
- Listed prices are not like for like: regions, minimum instance sizes, networking, storage, commitments and availability differ and are not priced here. Check the provider before relying on a figure.
Providers checked but not listed
- Google Cloud: The compute pricing pages (cloud.google.com/compute/gpus-pricing and vm-instance-pricing) showed no GPU prices when opened.
- RunPod: The pricing page shows two different sets of on-demand prices for the same GPUs (the visible cards and the page’s structured data), so no single listed price could be confirmed.
- Vast.ai: Its pricing page shows only relative claims ("50%+ cheaper"), no listed price. Marketplace prices change by offer.
- Voltage Park: Its pricing pages say "Contact for pricing" for every GPU; one homepage line says "Starting at $1.99/hr" without a GPU model or unit.
B200 rental prices: common questions
How much does it cost to rent a B200?
The lowest listed on-demand price among listings checked in the last 7 days is $6.00 per GPU-hour at Hyperstack, checked Oct 15, 2026. The table above lists every provider, variant and pricing type.
Which providers list B200 prices?
Nebius, Together AI, CoreWeave, Hyperstack, Lambda, Oracle Cloud, AWS.
Related: All GPU prices · Rental cost calculator · H100 · H200 · A100 80GB · L40S · MI300X
Cite this page
Last updated Oct 15, 2026. Figures may change as records are updated — include the date you retrieved them.
| APA | Data Center Census. (2026). NVIDIA B200 rental prices. Retrieved from https://datacentercensus.com/gpu-prices/b200 |
|---|---|
| MLA | “NVIDIA B200 rental prices.” Data Center Census, Oct 15, 2026, https://datacentercensus.com/gpu-prices/b200. |
| Link | https://datacentercensus.com/gpu-prices/b200 |
Republishing a figure or table? Attribute it to Data Center Census with a link to this page. The underlying data is downloadable as JSON and CSV; the press kit has standing figures and contact details.