High-end Blackwell GPU for large-scale AI training and inference.
Weekly median price per GPU per hour · Get the data
By provider shows one card per company, placed where its best offer ranked. By configuration lists every offer.
What you can rent comes first: in stock, then waitlist, then not reported, then out of stock. Priced offers rank ahead of quote-only, and on-demand ahead of other billing types.
Within a group, five factors set the order:
We match the provider, its country, GPU model, form factor, billing type, availability and instance name. Matching is partial and case-insensitive. Everyday words work too, so "interruptible" finds spot and "sold out" finds out of stock.
No providers match .
Heads up: A provider's own page may quote a different figure, for example on tax, region or a promotion. A monthly-only plan shows a derived hourly rate. Verify before provisioning. More on how we price.
One B200 has 180 GB of VRAM. In practice, that's enough memory for roughly 284B parameters at 4-bit or 76B at 16-bit, assuming a 32K context. Below are some open-weight LLMs, with the estimated memory and GPUs each one needs.
| Model | Memory (INT4 / FP4) | B200s needed | Cost /hr | Cost /mo |
|---|---|---|---|---|
Qwen3.8-27B
27B
|
18 GB
|
1
|
$6.79
|
$4,889
|
Gemma 4 31B
30.7B
|
22 GB
|
1
|
$6.79
|
$4,889
|
GPT-OSS-120B
117B · 5.1B active
|
67 GB
|
1
|
$6.79
|
$4,889
|
GLM-5.3-Flash
320B · 18B active
|
178 GB
|
2
|
$13.58
|
$9,778
|
MiniMax-M3
428B · 23B active
|
239 GB
|
2
|
$13.58
|
$9,778
|
DeepSeek V4.1 Flash
552B · 16B active
|
306 GB
|
2
|
$13.58
|
$9,778
|
Kimi K3
2.8T · 104B active
|
1,544 GB
|
10
|
–
|
–
|
Estimates based on the median on-demand rate. Memory is weights plus FP8 KV cache at 32K context per request (FP16/BF16 in the 16-bit column). GPU counts assume 90% of advertised VRAM is usable. No guarantee of runtime support, usable performance, or that a matching quantized build exists. Pricing methodology.
Nvidia B200 · Per GPU
| Compute · dense | |
|---|---|
| FP4 | 9,000 TFLOPS 18,000 with sparsity |
| FP8 | 4,500 TFLOPS 9,000 with sparsity |
| FP16 / BF16 | 2,250 TFLOPS 4,500 with sparsity |
| INT8 | 4,500 TOPS 9,000 with sparsity |
| FP32 | 75 TFLOPS |
| FP64 | 37 TFLOPS |
| Precision support | FP4FP6FP8FP16BF16TF32FP32FP64INT8 |
| Memory | |
|---|---|
| Capacity | 180 GB HBM3e |
| Bandwidth | 7,700 GB/s |
| ECC | Yes |
| Silicon | |
|---|---|
| Architecture | Blackwell |
| Process | TSMC 4NP |
| Transistors | 208 billion |
| Fabric and host | |
|---|---|
| GPU interconnect | NVLink 1,800 GB/s |
| Host interface | PCIe 5.0 x16 |
| Power | |
|---|---|
| Board power | 1,000 W |
| Cooling | Passive |
| Platform | |
|---|---|
| Partitioning | MIG, up to 7 instances |
Source: official Nvidia B200 datasheet.
As of September 29, 2026, the median on-demand price is $6.79 per GPU per hour across 19 providers with a priced on-demand config.
| Billing type | Configs | Median /GPU/hr | Cheapest |
|---|---|---|---|
On-demand | 64 | $6.79 | $3.75 (in stock, Packet·ai) |
Reserved | 57 | $5.62 | $3.75 |
Spot | 29 | $4.52 | $3.51 (in stock, Verda) |
10 providers also quote the B200 on a custom contract, priced per deal.
At 720 hours per month, one B200 costs an estimated $4,889 at the median on-demand price. Cheapest verified in stock: $2,700 per month on-demand, $3,787 reserved, $2,527 spot.
GetDeploying currently tracks B200 configs from 40 providers. The cheapest verified in-stock on-demand configs come from Packet·ai, Lium, GPU.ai and Verda. See the full price comparison above for every provider and config.
As of September 29, 2026, the median on-demand price has been flat over the past 90 days, at about $6.79 per GPU per hour, though it is about 17% above where it was a year ago.
One B200 runs models up to roughly 284B parameters at 4-bit quantization or 76B at 16-bit, assuming a 32K context. Larger models run across multiple GPUs: the model table above shows the estimated memory and GPU count for popular open-weight LLMs.
180GB HBM3e with 1800 GB/s NVLink. Blackwell FP4 Tensor Cores for significantly improved inference throughput over H100. Designed for training and serving the largest foundation models.
Nvidia H100
80 GB HBM3 · 56% less bandwidth than the B200
Nvidia B300
288 GB HBM3e · released Q1 2025
Nvidia H200
141 GB HBM3e · 38% less bandwidth than the B200
Last updated
