Skip to content
AuraMatrix

Compute

What a GPU costs per hour.

The raw material behind every token: renting one GPU for one hour across clouds, GPU clouds and marketplaces. Same rules as the token index: every number comes from a stored snapshot, and records are added, never edited.

Last sweep 2026-11-15 22:16 UTC9 providersrecording since 2026-11-11USD per GPU-hour

GPU-hour reference pricestrialMedian on-demand price for one GPU for one hour across providers. How it is built

Latest price per GPU. "On demand" is the cheapest list price for a GPU you can start now and keep; "lowest" includes spot, reserved and market prices.
GPUMemoryOn demand, fromOn demand, up toLowest of any kindMarketplace medianProviders
NVIDIA B300Blackwell Ultra generation, data centre268 GB9.99Together AI17.80AWS4.588CoreWeave · Spot—3
NVIDIA B200Blackwell generation, data centre180 GB5.98RunPod Community Cloud16.11Google Cloud4.09Together AI · Spot10.6318 asks8
NVIDIA H200Hopper generation, data centre141 GB3.59RunPod Community Cloud10.60Microsoft Azure2.58CoreWeave · Spot4.84118 asks8
NVIDIA H100Hopper generation, data centre80 GB1.99RunPod Community Cloud12.29Microsoft Azure1.418Microsoft Azure · Spot1.86821 asks9
NVIDIA A100 80GBAmpere generation, data centre80 GB1.19RunPod Community Cloud5.028Google Cloud0.679Microsoft Azure · Spot0.84435 asks7
NVIDIA L40SData centre, inference and graphics48 GB0.79RunPod Community Cloud3.766AWS0.774Vast.ai · Market (median ask)0.77434 asks5
NVIDIA RTX 5090Consumer card, rented on GPU clouds and marketplaces32 GB0.69RunPod Community Cloud1.39RunPod Secure Cloud0.584Vast.ai · Market (median ask)0.58464 asks3
NVIDIA RTX 4090Consumer card, rented on GPU clouds and marketplaces24 GB0.34RunPod Community Cloud0.89RunPod Secure Cloud0.34RunPod Community Cloud · On demand0.50164 asks3
NVIDIA V100 32GBVolta generation, data centre32 GB3.901AWS—0.39AWS · Spot—1
NVIDIA T4Turing generation, inference16 GB0.526AWS0.978AWS0.209AWS · Spot—1

In mainland China, in CNY

China’s clouds sell a different shelf: the newest NVIDIA cards cannot be exported there, so the same money buys older models. These are list prices in yuan, shown as published and never converted, so they stand next to the dollar prices rather than among them.

CNY per GPU-hour, pay as you go. Last sweep 2026-11-15 22:19 UTC.
GPUMemoryOn demand, fromOn demand, up toStock at the lowest price
NVIDIA L20Data centre, a model made for the China market48 GB¥14.41Alibaba Cloud · ebmgn8is¥23.94none in stock
NVIDIA A10Ampere generation, inference and graphics24 GB¥9.53Alibaba Cloud · gn7i¥13.312 of 4 zones
NVIDIA V100 32GBVolta generation, data centre32 GB¥19.74Alibaba Cloud · gn6e—5 of 5 zones
NVIDIA V100Volta generation, data centre16 GB¥26.46Alibaba Cloud · gn6v—none in stock
NVIDIA T4Turing generation, inference16 GB¥11.63Alibaba Cloud · gn6i¥17.192 of 6 zones

Same GPU, different providers

How to read this

Each dot is one provider at its cheapest on-demand price for one GPU-hour. The line runs from zero to the dearest provider for that GPU, so the gaps between dots are to scale within a row.

  • B2002.7×$5.98 RunPod Community CloudGoogle Cloud $16.11
  • H2003.0×$3.59 RunPod Community CloudMicrosoft Azure $10.60
  • H1005.5×$1.99 RunPod Community CloudGoogle Cloud $10.98
  • A100 80GB4.2×$1.19 RunPod Community CloudGoogle Cloud $5.028
  • L40S2.8×$0.79 RunPod Community CloudCoreWeave $2.25

We are building an exchange for compute. What it will be, and the waitlist →

How to read these prices

One GPU, one hour
Every price is in US dollars per GPU-hour. When a cloud sells whole machines, we divide the price by the number of GPUs, so that figure includes the bundled CPUs, memory and disk. A GPU cloud or marketplace usually pairs the GPU with a smaller machine. The two are comparable, not identical.
Kinds of price
On demand: start now, stop any time. Spot: cheaper, but the provider can take the machine back. Reserved: a commitment of one or three years, shown per hour. Market: the median of what hosts on a marketplace were asking when we looked; it moves all the time and hosts differ in reliability.
Where it comes from
Public price lists and APIs, fetched once a day and stored with a hash. We do not buy or measure anything: these are asking prices, not invoices, and we do not check availability.
What is not here yet
Google Cloud, Together and the Chinese clouds are being added. Contract prices and negotiated discounts are private and not shown.

The whole dataset as JSON. Corrections: hello@auramatrix.ai