NVIDIA A100 rental pricing
Ampere architecture · 40 to 80 GB HBM2e memory · 2.0 TB/s bandwidth · 250 to 400 W board power · Released 2020
Price history
Weekly median price per GPU per hour
At a glance
prices checked just now- Cheapest in stock
- $0.43/hr
- Community · 2× A100 SXM4 · Vast.ai
- Median price
- $1.55/hr
- On-demand · 24 providers
- 90-day trend
- Too early
- Last 30 days · fills in as prices are checked
- Coverage
- 28 providers
- 90 price points · 4 variants
- $ per FP16 PFLOP-hour
- $2.85
- Rank 3 of 19 priced GPUs · lower is better
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 | |
|---|---|---|---|---|
| 2 GPUs Community | $0.43 | just now | View → | |
 GPU.ai | 1 GPU Community | $0.54 | just now | View → |
| 1 GPU Spot | $0.64 | just now | View → | |
 Civo | 1 GPU Reserved | $0.69 | just now | View → |
| GPU.ai | 1 GPU | $0.89 | just now | View → |
 Oblivus | 1 GPU Reserved | $0.95 | just now | View → |
| 1 GPU Community | $1.00 | just now | View → | |
| 1 GPU Spot | $1.08 | just now | View → | |
 Thunder Compute | 1 GPU | $1.09 | just now | View → |
| Civo | 1 GPU | $1.09 | just now | View → |
 Spheron | 1 GPU Spot | $1.12 | just now | View → |
 Lium | 1 GPU Community | $1.12 | just now | View → |
| 8 GPUs Spot | $1.21 | just now | View → | |
| 1 GPU | $1.28 | just now | View → | |
| 1 GPU | $1.35 | just now | View → | |
 Beam | 1 GPU | $1.36 | just now | View → |
 Shadeform Denvr Dataworks capacity | 8 GPUs | $1.40 | just now | View → |
| Spheron | 1 GPU | $1.43 | just now | View → |
 Gcore | 1 GPU | $1.44 | just now | View → |
| Oblivus | 1 GPU | $1.47 | just now | View → |
| 1 GPU | $1.49 | just now | View → | |
| 1 GPU | $1.50 | just now | View → | |
| 1 GPU | $1.59 | just now | View → | |
 Lyceum | 1 GPU | $1.59 | just now | View → |
 Koyeb | 1 GPU | $1.60 | just now | View → |
| 1 GPU | $1.99 | just now | View → | |
 Cerebrium | 1 GPU | $2.00 | just now | View → |
| 1 GPU | $2.00 | just now | View → | |
| 1 GPU | $2.10 | just now | View → | |
 Cirrascale | 8 GPUs Reserved | $2.30 | just now | View → |
| 8 GPUs | $2.70 | just now | View → | |
| 8 GPUs | $2.74 | just now | View → | |
| 8 GPUs | $3.40 | just now | View → | |
| 1 GPU | $4.00 | just now | View → | |
 Replicate | 1 GPU | $5.04 | just now | View → |
| Replicate | 4 GPUs Reserved | $5.04 | just now | View → |
| 1 GPU | $1.81 | 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 A100 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 | 1A100 SXM4 | $0.89 |
| Gemma 3 27B Google | 27.4B | 26 GiB | 1A100 SXM4 | $0.89 |
| Qwen3-32B Alibaba | 32.8B | 32 GiB | 1A100 SXM4 | $0.89 |
| Llama 3.3 70B Instruct Meta | 70.6B | 67 GiB | 1A100 SXM4 | $1.07 |
| gpt-oss-120b OpenAI | 117B· 5.1B active | 109 GiB | 2A100 SXM4 | $2.14 |
| Qwen3-235B-A22B Alibaba | 235B· 22.0B active | 220 GiB | 4A100 SXM4 | $4.28 |
| DeepSeek-R1 DeepSeek | 684B· 37.0B active | 638 GiB | 16A100 SXM4 | $17.12 |
| 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 A100 pricing questions
How much does it cost to rent a NVIDIA A100 per hour?
The cheapest NVIDIA A100 in stock right now is $0.43 per GPU-hour community from Vast.ai for the 2× A100 SXM4 configuration. The median on-demand rate across 24 providers is $1.55 per GPU-hour.
How much does a NVIDIA A100 cost per month?
Running one NVIDIA A100 around the clock for a month (730 hours) costs about $312 at the cheapest in-stock rate of $0.43 per hour. At the median on-demand rate it is about $1,128 a month. Reserved terms usually come in below that; ask the provider for a quote.
Which provider has the cheapest NVIDIA A100?
Vast.ai has the cheapest NVIDIA A100 in stock right now at $0.43 per GPU-hour community for the 2× A100 SXM4 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 A100 rental prices going up or down?
We have 5 weeks of price history for the NVIDIA A100, not yet enough like-for-like weeks to call a trend: the set of providers we track is still growing, and a median over more sellers is not a price move. The chart above fills in as our daily checks accumulate.
What is the difference between on-demand, spot and reserved NVIDIA A100 pricing?
On-demand is pay-as-you-go with no commitment. Spot is spare capacity sold at a discount that the provider can reclaim at short notice. Reserved is a committed term, usually months, in exchange for a lower rate. Community is capacity from individual hosts on a marketplace, cheaper and less uniform. This week's median rates for the NVIDIA A100 are on-demand $1.59 (25 providers), spot $1.14 (4 providers), reserved $1.63 (4 providers), community $0.85 (4 providers), per GPU-hour.
How many cloud providers rent the NVIDIA A100?
We track 28 providers with 90 live prices for the NVIDIA A100 across 4 variants. Sold-out listings stay in the table so you can see who normally carries it, but never set the cheapest figure.
Which NVIDIA A100 variant is cheapest to rent?
The SXM4 is the cheapest NVIDIA A100 variant right now at $0.43 per GPU-hour. Cheapest in-stock rate per variant: SXM4 from $0.43, PCIe from $0.43, PCIe from $0.87, SXM4 from $0.88. The variants differ in memory, bandwidth and interconnect, so the cheapest card is not always the cheapest job.
What LLMs can you run on one NVIDIA A100?
With 8-bit weights and a single sequence of context, one NVIDIA A100 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.
How does the NVIDIA A100 compare on price per PFLOP-hour?
At its cheapest in-stock on-demand rate the NVIDIA A100 delivers dense FP16 compute at $2.85 per PFLOP-hour, which ranks 3 of 19 priced GPUs we cover on that basis. Lower is better: it is the rental price divided by peak FP16 throughput.
Where does this NVIDIA A100 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.