The standard data center 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 H100 has 80 GB of VRAM on the SXM variant. In practice, that's enough memory for roughly 120B parameters at 4-bit or 31B 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) | H100s needed | Cost /hr | Cost /mo |
|---|---|---|---|---|
Qwen3.8-27B
27B
|
18 GB
|
1
|
$3.38
|
$2,434
|
Gemma 4 31B
30.7B
|
22 GB
|
1
|
$3.38
|
$2,434
|
GPT-OSS-120B
117B · 5.1B active
|
67 GB
|
1
|
$3.38
|
$2,434
|
GLM-5.3-Flash
320B · 18B active
|
178 GB
|
4
|
$13.52
|
$9,734
|
MiniMax-M3
428B · 23B active
|
239 GB
|
4
|
$13.52
|
$9,734
|
DeepSeek V4.1 Flash
552B · 16B active
|
306 GB
|
8
|
$27.04
|
$19,469
|
Kimi K3
2.8T · 104B active
|
1,544 GB
|
22
|
–
|
–
|
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 H100 · Per GPU · SXM
| Compute · dense | |
|---|---|
| FP8 | 1,979 TFLOPS 3,958 with sparsity |
| FP16 / BF16 | 989.5 TFLOPS 1,979 with sparsity |
| INT8 | 1,979 TOPS 3,958 with sparsity |
| FP32 | 67 TFLOPS |
| FP64 | 34 TFLOPS |
| Precision support | FP8FP16BF16TF32FP32FP64INT8 |
| Memory | |
|---|---|
| Capacity | 80 GB HBM3 |
| Bandwidth | 3,350 GB/s |
| Bus width | 5,120-bit |
| ECC | Yes |
| Silicon | |
|---|---|
| Architecture | Hopper |
| Process | TSMC 4N |
| Transistors | 80 billion |
| Shader cores | 16,896 CUDA cores |
| Matrix cores | 528 Tensor cores |
| Compute units | 132 SMs |
| Fabric and host | |
|---|---|
| GPU interconnect | NVLink 900 GB/s |
| Host interface | PCIe 5.0 x16 |
| Power | |
|---|---|
| Board power | 700 W |
| Platform | |
|---|---|
| Partitioning | MIG, up to 7 instances |
Source: official Nvidia H100 datasheet.
The H100 ships in 3 versions with different memory, bandwidth or interconnect. The figures at the top of this page are for the H100 SXM.
| Variant | Memory | Memory bandwidth | Interconnect |
|---|---|---|---|
H100 SXM | 80 GB | 3,350 GB/s | NVLink 900 GB/s |
H100 PCIe | 80 GB | 2,000 GB/s | NVLink 600 GB/s |
H100 NVL | 94 GB | 3,938 GB/s | NVLink 600 GB/s |
As of October 2, 2026, the median on-demand price is $3.38 per GPU per hour across 40 providers with a priced on-demand config.
| Billing type | Configs | Median /GPU/hr | Cheapest |
|---|---|---|---|
On-demand | 227 | $3.38 | $1.30 (in stock, Lium) |
Reserved | 174 | $3.09 | $1.53 (1 mo, in stock, HyperAI) |
Spot | 60 | $2.23 | $0.67 (in stock, Vast.ai) |
5 providers also quote the H100 on a custom contract, priced per deal.
At 720 hours per month, one H100 costs an estimated $2,434 at the median on-demand price. Cheapest verified in stock: $936 per month on-demand, $1,102 reserved (1 mo), $482 spot.
GetDeploying currently tracks H100 configs from 57 providers. The cheapest verified in-stock on-demand configs come from Lium, Vast.ai, HyperAI and Hyperstack. See the full price comparison above for every provider and config.
As of October 2, 2026, the median on-demand price has been flat over the past 90 days, at about $3.38 per GPU per hour, though it is about 12% above where it was a year ago.
One H100 runs models up to roughly 120B parameters at 4-bit quantization or 31B 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.
80GB HBM3 with 900 GB/s NVLink for multi-node training. FP8 Transformer Engine for improved training throughput over A100. The standard choice for large-scale AI training and high-throughput inference.
Nvidia B300
288 GB HBM3e · 2.4x the bandwidth of the H100
Nvidia H200
141 GB HBM3e · 43% more bandwidth than the H100
Nvidia RTX 5090
32 GB GDDR7 · 47% less bandwidth than the H100
Nvidia A100
80 GB HBM2e · 42% less bandwidth than the H100
AMD MI300X
192 GB HBM3 · 58% more bandwidth than the H100
Last updated
