Loading prices…
Lowest on-demand
$0.40per GPU·h · ≈ $293/month
at Vast.aiMedian on-demand
$1.72/GPU·h
across 6 providers
Lowest spot
—
Reserved from $0.69
FP16/BF16 cost
$1.29/PFLOP·h
#5 of 72 GPUs for value
Providers
6
33 configurations
Price history
Daily A100 PCIe 40GB prices per GPU-hour. The shaded period is reconstructed from public archives: the version history of SkyPilot's price catalogs, Internet Archive copies of providers' pricing pages and the open gpu-price-feed project (fewer providers than live data). After that, PetaPrice's own tracking.
Lowest on-demandMedian on-demandLowest spot
Price by provider
Cheapest On-demand A100 PCIe 40GB offer at each provider, per GPU-hour. Shorter and greener is cheaper.
Vast.ai🇺🇸$0.40best
Civo🇬🇧$1.092.7×
Crusoe🇺🇸$1.453.6×
Lambda🇺🇸$1.995.0×
Paperspace🇺🇸$3.117.7×
Sesterce🇫🇷$3.518.7×
Thin vertical line = median across providers ($1.45).
Every provider we track
Pricing
Size
Host
Provider HQ
6 providers · 33 configurations
Sort
- Vast.aicommunityOn-demand🇺🇸 HQ in United Statesvia Vast.aifrom · On-demandCheapest$0.40/GPU·hCheapest
- 2× GPU · PCIe24 vCPU · 189 GB RAM · 853 GB disk$0.40 /GPU·h$0.80 /hOn-demand🇺🇸 California, US● availableRent ↗
- CivoOn-demandReserved🇬🇧 HQ in United Kingdomvia Civo pricing pagefrom · On-demand2.7×$1.09/GPU·h2.7×
- CrusoeOn-demand🇺🇸 HQ in United Statesvia Shadeformfrom · On-demand3.6×$1.45/GPU·h3.6×
- 1× GPU · PCIe12 vCPU · 120 GB RAM · 1,088 GB disk$1.45 /GPU·h$1.45 /hOn-demand🇺🇸 US, Fairview, OH○ sold outRent ↗
- 8× GPU · PCIe96 vCPU · 960 GB RAM · 7,808 GB disk$1.45 /GPU·h$11.6 /hOn-demand🇺🇸 US, Fairview, OH○ sold outRent ↗
- 4× GPU · PCIe48 vCPU · 480 GB RAM · 2,048 GB disk$1.45 /GPU·h$5.80 /hOn-demand🇺🇸 US, Fairview, OH○ sold outRent ↗
- LambdaOn-demand🇺🇸 HQ in United Statesvia Lambda pricing pagefrom · On-demand5.0×$1.99/GPU·h5.0×
- PaperspaceOn-demand🇺🇸 HQ in United Statesvia Shadeformfrom · On-demand7.7×$3.11/GPU·h7.7×
- 8× GPU · NVLink96 vCPU · 720 GB RAM · 2,000 GB disk$3.11 /GPU·h$24.9 /hOn-demand🇳🇱 NL, Amsterdam○ sold outRent ↗
- 4× GPU · NVLink48 vCPU · 360 GB RAM · 1,000 GB disk$3.11 /GPU·h$12.4 /hOn-demand🇳🇱 NL, Amsterdam○ sold outRent ↗
- 2× GPU · NVLink24 vCPU · 180 GB RAM · 500 GB disk$3.13 /GPU·h$6.26 /hOn-demand🇳🇱 NL, Amsterdam○ sold outRent ↗
- SesterceOn-demand🇫🇷 HQ in Francevia Sestercefrom · On-demand8.7×$3.51/GPU·h8.7×
- 1× GPU · PCIe8 vCPU · 90 GB RAM · 1,000 GB disk$3.51 /GPU·h$3.51 /hOn-demandUSA-NewYork-1● availableRent ↗
What can the A100 PCIe 40GB run?
Memory for the weights plus one 32K-token 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 | $/mo |
|---|---|---|---|---|---|
gpt-oss-20b OpenAI · 20.9B | 20.9B · 3.6B active | 20 GiB | 1 | $0.40 at Vast.ai | $293 |
Gemma 3 27B Google · 27.4B | 27.4B | 28 GiB | 1 | $0.40 at Vast.ai | $293 |
Qwen3-32B Alibaba · 32.8B | 32.8B | 39 GiB | 2 | $0.80 at Vast.ai | $586 |
Llama 3.3 70B Instruct Meta · 70.6B | 70.6B | 76 GiB | 4 | $1.61 at Vast.ai | $1,172 |
gpt-oss-120b OpenAI · 117B | 117B · 5.1B active | 110 GiB | 4 | $1.61 at Vast.ai | $1,172 |
Qwen3-235B-A22B Alibaba · 235B | 235B · 22B active | 225 GiB | 8 | $3.21 at Vast.ai | $2,345 |
DeepSeek-R1 DeepSeek · 684B | 684B · 37B active | 640 GiB | 32 | $12.85 at Vast.ai | $9,379 |
Kimi K3 Moonshot AI · 2.78T | 2.78T · 104B active | 2.53 TiB | 64+ | — | — |
Assumes 90% of each GPU's memory is usable and rounds up to 1, 2, 4, 8… GPUs for tensor parallelism (up to 64). $/hr is the GPU count times the cheapest in-stock on-demand rate per GPU; check that the provider rents instances that large. Serving many requests at once needs more memory for context.
Alternatives to the A100 PCIe 40GB
GPUs with similar compute and memory, and their lowest on-demand price right now.
- A100 SXM 40GB40 GB · 0.312 PF FP16$0.41/h$1.31/PFLOP·h
- A100 SXM 80GB80 GB · 0.312 PF FP16$1.00/h$3.20/PFLOP·h
- A100 PCIe 80GB80 GB · 0.312 PF FP16$1.11/h$3.56/PFLOP·h
- L40S48 GB · 0.362 PF FP16$0.38/h$1.05/PFLOP·h
Trainium32 GB · 0.19 PF FP16$1.34/h$7.07/PFLOP·h- Inferentia232 GB · 0.19 PF FP16$0.76/h$3.99/PFLOP·h
A100 PCIe 40GB pricing questions
How much does it cost to rent an A100 PCIe 40GB?
On-demand A100 PCIe 40GB prices start at $0.40 per GPU-hour (Vast.ai). The median across 6 providers is $1.72, and the most expensive is $3.51. Running one GPU around the clock costs about $293 per month at the lowest price.
How much compute do you get per dollar with the A100 PCIe 40GB?
At $0.40 per hour and 0.312 PF of dense FP16/BF16, the A100 PCIe 40GB costs $1.29 per PFLOP-hour. That ranks 5 of 72 GPUs with on-demand prices on PetaPrice. Figures are datasheet peaks; real utilization is typically 30–60%.
What are the A100 PCIe 40GB's key specs?
40 GB HBM2 with 1,555 GB/s of memory bandwidth, Ampere architecture (2020), PCIe form factor, 250 W. Dense peak: 0.312 PF FP16 / BF16, 0.624 PF INT8, 0.156 PF TF32, 0.0195 PF FP32, 0.0195 PF FP64.
How current are these prices?
Prices were collected on Nov 10, 2026, 17:01 UTC from provider APIs and public catalogs and refresh about every 2 hours. They are indicative list prices; always confirm on the provider's website before booking.