Skip to content
Get my AI grade

Rank.ai original data

GPU rental price index

A single index for open-market GPU rental pricing: the median of the median hourly prices across every GPU configuration listed on the Akash marketplace.
median across GPU configurations$0.20 / GPU-hr

Behind the number

The full distribution

Open the board →
  • NVIDIA M4000 8 GB PCIe$0.04
  • NVIDIA P4 8 GB PCIe$0.06
  • NVIDIA RTX 3060TI 8 GB PCIe$0.14
  • NVIDIA GTX1050 2 GB PCIe$0.14
  • NVIDIA RTX 3090 24 GB PCIe$0.15
  • NVIDIA P40 24 GB PCIe$0.15
  • NVIDIA T4 16 GB PCIe$0.16
  • NVIDIA A10G 24 GB PCIe$0.18
  • NVIDIA V100 32 GB PCIe$0.21
  • NVIDIA V100 16 GB SXM2$0.25
  • NVIDIA RTX 4000ADA 20 GB PCIe$0.29
  • NVIDIA RTX 4090 24 GB PCIe$0.34

The board in rank order. The headline number is the middle of this distribution, highlighted, so the spread around it is what the index smooths. Showing the top 12 of 16 ranked entries.

median across GPU configurations by entry
RankEntrymedian across GPU configurations
1NVIDIA M4000 8 GB PCIe$0.04 / GPU-hr
2NVIDIA P4 8 GB PCIe$0.06 / GPU-hr
3NVIDIA RTX 3060TI 8 GB PCIe$0.14 / GPU-hr
4NVIDIA GTX1050 2 GB PCIe$0.14 / GPU-hr
5NVIDIA RTX 3090 24 GB PCIe$0.15 / GPU-hr
6NVIDIA P40 24 GB PCIe$0.15 / GPU-hr
7NVIDIA T4 16 GB PCIe$0.16 / GPU-hr
8NVIDIA A10G 24 GB PCIe$0.18 / GPU-hr
9NVIDIA V100 32 GB PCIe$0.21 / GPU-hr
10NVIDIA V100 16 GB SXM2$0.25 / GPU-hr
11NVIDIA RTX 4000ADA 20 GB PCIe$0.29 / GPU-hr
12NVIDIA RTX 4090 24 GB PCIe$0.34 / GPU-hr

Over time

How the number has moved

$0.18$0.21$0.24$0.26Oct 26Oct 28Oct 30
View chart data
median across GPU configurations by publication
As ofmedian across GPU configurations
Oct 30, 05:00 UTC$0.20 / GPU-hr
Oct 30, 04:00 UTC$0.20 / GPU-hr
Oct 30, 03:00 UTC$0.20 / GPU-hr
Oct 29, 23:00 UTC$0.20 / GPU-hr
Oct 29, 19:00 UTC$0.20 / GPU-hr
Oct 29, 15:00 UTC$0.20 / GPU-hr
Oct 29, 11:00 UTC$0.21 / GPU-hr
Oct 29, 07:00 UTC$0.21 / GPU-hr
Oct 29, 03:00 UTC$0.21 / GPU-hr
Oct 28, 23:00 UTC$0.25 / GPU-hr
Oct 28, 19:00 UTC$0.22 / GPU-hr
Oct 28, 15:00 UTC$0.22 / GPU-hr
Oct 28, 11:00 UTC$0.17 / GPU-hr
Oct 28, 07:00 UTC$0.19 / GPU-hr
Oct 28, 03:00 UTC$0.19 / GPU-hr
Oct 27, 23:00 UTC$0.19 / GPU-hr
Oct 27, 19:00 UTC$0.18 / GPU-hr
Oct 27, 15:00 UTC$0.19 / GPU-hr
Oct 27, 11:00 UTC$0.17 / GPU-hr
Oct 27, 07:00 UTC$0.18 / GPU-hr
Oct 27, 03:00 UTC$0.19 / GPU-hr
Oct 26, 23:00 UTC$0.22 / GPU-hr
Oct 26, 19:00 UTC$0.20 / GPU-hr
Oct 26, 15:00 UTC$0.21 / GPU-hr
Oct 26, 11:00 UTC$0.21 / GPU-hr
Oct 26, 07:00 UTC$0.21 / GPU-hr
Oct 26, 03:00 UTC$0.21 / GPU-hr

About this metric

No single GPU tells the whole rental market's story. This index compresses the open marketplace into one number by taking the median configuration's median hourly price across everything listed, from consumer RTX cards through data-center H200s. It answers a plain question: what did an hour of GPU compute cost on the open market at the latest snapshot?

Each ranked configuration on the underlying board carries its own median price per GPU-hour, computed from active positive-price listings. The index is the middle value of those medians, so one scarce flagship or one dumped consumer card cannot drag it around. The construction is deterministic and recomputed at every published snapshot.

Hardware generations, marketplace supply and demand cycles all show up in aggregate GPU pricing. Read it alongside the H100 line: when the flagship and the index diverge, the market is repricing the newest hardware rather than compute in general.

Charts

Related charts

All charts

See how often AI names you.

Enter your website. In about two minutes, Rank.ai asks ChatGPT, Claude and Gemini 12 questions your customers ask and grades how often they name you.

  • Your grade out of 100How often AI names you, cites your site, and how high it ranks you.
  • Who gets namedEvery competitor in the answers, most named first.
  • The pages AI readsThe sources behind each answer.
  • Three fixesWhat to fix first, with a brief for the first article.