NVIDIA H100 NVL rental pricing
Hopper architecture · 94 GB HBM3 memory · 3.9 TB/s bandwidth · 400 W board power · Released 2023
Price history
Weekly median price per GPU per hour
At a glance
prices checked just now- Cheapest in stock
- $2.37/hr
- Community · 1× H100 NVL · GPU.ai
- Median price
- $3.19/hr
- On-demand · 3 providers
- 90-day trend
- Down 37.3%
- Weekly median on-demand, 4 weeks
- Coverage
- 5 providers
- 9 price points
Every provider we track
Cheapest listing per provider and pricing type. Sold-out rows stay so you can see who normally carries it.
| Provider | Configuration | Price/GPU-hr | Checked | |
|---|---|---|---|---|
 GPU.ai | 1× H100 NVL Community | $2.37 | just now | View → |
| 1× H100 NVL Community | $2.59 | just now | View → | |
| 2× H100 NVL Community | $2.67 | just now | View → | |
| 1× H100 NVL | $3.19 | just now | View → | |
| GPU.ai | 1× H100 NVL | $3.19 | just now | View → |
| 1× H100 NVL | $6.98 | just now | View → | |
 Shadeform Massed Compute capacity | 8× H100 NVL | $3.16 | just now | View → |
Per GPU-hour. Spot, reserved and community pricing differ from on-demand; click through for the provider's current figure.
What can the NVIDIA H100 NVL run?
Memory for the weights plus one sequence of context, how many GPUs hold it, and what that replica costs at the cheapest in-stock on-demand rate.
| Model | Params | Memory | GPUs | $/hr |
|---|---|---|---|---|
| gpt-oss-20b OpenAI | 20.9B· 3.6B active | 20 GiB | 1 | $3.19 |
| Gemma 3 27B Google | 27.4B | 26 GiB | 1 | $3.19 |
| Qwen3-32B Alibaba | 32.8B | 32 GiB | 1 | $3.19 |
| Llama 3.3 70B Instruct Meta | 70.6B | 67 GiB | 1 | $3.19 |
| gpt-oss-120b OpenAI | 117B· 5.1B active | 109 GiB | 2 | $6.38 |
| Qwen3-235B-A22B Alibaba | 235B· 22.0B active | 220 GiB | 4 | $12.76 |
| DeepSeek-R1 DeepSeek | 684B· 37.0B active | 638 GiB | 8 | $25.52 |
| Kimi K3 Moonshot AI | 2.78T· 104B active | 2.53 TiB | Over 16 | — |
Sized for one sequence at 4K context with 16-bit KV cache and 10% of memory held back. Serving many users at once needs more; the LLM GPU calculator sizes for concurrency and throughput.
NVIDIA H100 NVL pricing questions
How much does it cost to rent a NVIDIA H100 NVL per hour?
The cheapest NVIDIA H100 NVL in stock right now is $2.37 per GPU-hour community from GPU.ai for the 1× H100 NVL configuration. The median on-demand rate across 3 providers is $3.19 per GPU-hour.
How much does a NVIDIA H100 NVL cost per month?
Running one NVIDIA H100 NVL around the clock for a month (730 hours) costs about $1,730 at the cheapest in-stock rate of $2.37 per hour. At the median on-demand rate it is about $2,329 a month. Reserved terms usually come in below that; ask the provider for a quote.
Which provider has the cheapest NVIDIA H100 NVL?
GPU.ai has the cheapest NVIDIA H100 NVL in stock right now at $2.37 per GPU-hour community for the 1× H100 NVL configuration. That is a community rate; see the provider table for the cheapest on-demand listing. Rates change often, so check the provider table for today's figure before you commit.
Are NVIDIA H100 NVL rental prices going up or down?
The weekly median on-demand rate for the NVIDIA H100 NVL is down 37.3% over the last 4 weeks, from $5.09 to $3.19 per GPU-hour.
What is the difference between on-demand, spot and reserved NVIDIA H100 NVL pricing?
On-demand is pay-as-you-go with no commitment. Community is capacity from individual hosts on a marketplace, cheaper and less uniform. This week's median rates for the NVIDIA H100 NVL are on-demand $3.19 (4 providers), community $2.59 (3 providers), per GPU-hour.
How many cloud providers rent the NVIDIA H100 NVL?
We track 5 providers with 9 live prices for the NVIDIA H100 NVL. Sold-out listings stay in the table so you can see who normally carries it, but never set the cheapest figure.
What LLMs can you run on one NVIDIA H100 NVL?
With 8-bit weights and a single sequence of context, one NVIDIA H100 NVL holds gpt-oss-20b (20.9B), Gemma 3 27B (27.4B), Qwen3-32B (32.8B), Llama 3.3 70B Instruct (70.6B). See the model fit table for the full ladder.
Where does this NVIDIA H100 NVL pricing data come from?
Prices are read daily from each provider's public price list or API and recorded whenever they change. Listings are shown per GPU-hour on-demand unless marked otherwise, and the chart covers the last 30 days. The newest listing was checked just now.