NVIDIA A40 48GB rental pricing
Ampere architecture · 48 GB GDDR6 memory · 0.7 TB/s bandwidth · 300 W board power · Released 2021
Price history
Weekly median price per GPU per hour
At a glance
prices checked 1h ago- Cheapest in stock
- $0.35/hr
- Community · 1× A40 48GB · RunPod
- Median price
- $0.49/hr
- On-demand · 1 provider
- 90-day trend
- Flat
- Weekly median on-demand, 4 weeks
- Coverage
- 2 providers
- 3 price points
- $ per FP16 PFLOP-hour
- $3.27
- 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.
Per GPU-hour. Spot, reserved and community pricing differ from on-demand; click through for the provider's current figure.
What can the NVIDIA A40 48GB 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 | $0.49 |
| Gemma 3 27B Google | 27.4B | 26 GiB | 1 | $0.49 |
| Qwen3-32B Alibaba | 32.8B | 32 GiB | 1 | $0.49 |
| Llama 3.3 70B Instruct Meta | 70.6B | 67 GiB | 2 | $0.98 |
| gpt-oss-120b OpenAI | 117B· 5.1B active | 109 GiB | 4 | $1.96 |
| Qwen3-235B-A22B Alibaba | 235B· 22.0B active | 220 GiB | 8 | $3.92 |
| DeepSeek-R1 DeepSeek | 684B· 37.0B active | 638 GiB | 16 | $7.84 |
| 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 A40 48GB pricing questions
How much does it cost to rent a NVIDIA A40 48GB per hour?
The cheapest NVIDIA A40 48GB in stock right now is $0.35 per GPU-hour community from RunPod for the 1× A40 48GB configuration. The median on-demand rate across 1 provider is $0.49 per GPU-hour.
How much does a NVIDIA A40 48GB cost per month?
Running one NVIDIA A40 48GB around the clock for a month (730 hours) costs about $255 at the cheapest in-stock rate of $0.35 per hour. At the median on-demand rate it is about $358 a month. Reserved terms usually come in below that; ask the provider for a quote.
Which provider has the cheapest NVIDIA A40 48GB?
RunPod has the cheapest NVIDIA A40 48GB in stock right now at $0.35 per GPU-hour community for the 1× A40 48GB 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 A40 48GB rental prices going up or down?
The weekly median on-demand rate for the NVIDIA A40 48GB has been flat, moving 0.0% over the last 4 weeks, from $0.49 to $0.49 per GPU-hour.
What is the difference between on-demand, spot and reserved NVIDIA A40 48GB 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 A40 48GB are on-demand $0.49 (1 provider), community $0.38 (2 providers), per GPU-hour.
How many cloud providers rent the NVIDIA A40 48GB?
We track 2 providers with 3 live prices for the NVIDIA A40 48GB. 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 A40 48GB?
With 8-bit weights and a single sequence of context, one NVIDIA A40 48GB holds gpt-oss-20b (20.9B), Gemma 3 27B (27.4B), Qwen3-32B (32.8B). See the model fit table for the full ladder.
How does the NVIDIA A40 48GB compare on price per PFLOP-hour?
At its cheapest in-stock on-demand rate the NVIDIA A40 48GB delivers dense FP16 compute at $3.27 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 A40 48GB 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 1h ago.