GPU EconomyCloud GPU prices, read every hour
Latest read 97 providers tracked14,205 prices read in 24 h

NVIDIA H100 SXM — what a GPU-hour costs

$3.53Price Index: H100 SXM on-demand median, 34 providers

Listed rates, not availability. Read . · Renting for a year or more? Reserved quotes ↓

H100 SXM · like-for-like price index, Sep 2025 = 100+19.4% since Sep 2025
100110120
119.4
Sep 2025JanMaySep 2026
same providers and regions both months, public catalog; the $ above is the Price Indexweekly history →
Cheapest$1.79Cudo Compute — 49% below the median
Spread6.2×$1.79 to $11.06 for the same card
Hyper premium+207%$9.90 vs $3.23 at specialists
Past 12 mo+19%like-for-like median, September 2025 = 100
TREND

In the last closed month (September 2026) the H100 SXM like-for-like median rose 2.6%, the second month running. Over 12 months, +19%.

Size your workload

Who is cheapest today, and what a month costs.

GPUs
Hours per month
Region
A month at the top pick$10,454Cudo Compute · $1.79 per GPU-hour, on-demand · no-kristiansand-1, GPU only · captured $10,162 less than at the Price Index

H100 SXM by provider

One table per market, on-demand first; cheapest first within each.

Show
On-demand · low risk — 39 rates (1 not confirmed, out of every figure), cheapest first · Price Index $3.53
Provider · what is sold$ / GPU-hour30 daysReadSource
Cudo ComputeGPU onlyAPIno-kristiansand-1$1.79−49% vs indexflatcaptured Deploy ↗
TensorPoolGPU only$1.99−44% vs indexstoppedlast confirmed · not in the figures
Seeweb€1.89 listed · ECB 1.1193 on October 23, 2026$2.12−40% vs indexflatcaptured Deploy ↗
Bitdeer AIfrom$4.72/node ÷ 2$2.36−33% vs indexflat · 13 dcaptured Deploy ↗
HPC-AI.COMfrom7d +10.0%$2.42−31% vs index+10.0% · 10 dcaptured Deploy ↗
Denvr Dataworks$2.45−31% vs indexflatcaptured Deploy ↗
Taiga Cloud (Northern Data Group)from$2.60−26% vs indexflat · 24 dcaptured Deploy ↗
Atlas Cloud$2.95−16% vs indexflat · 27 dcaptured Deploy ↗
VESSL AI$2.98−16% vs indexflat · 12 dcaptured Deploy ↗
Civo$2.99−15% vs indexflat · 27 dcaptured Deploy ↗
FPT AI FactorySoutheast Asia$2.99−15% vs indexflat · 11 dcaptured Deploy ↗
HexGridSecure Cloud$2.99−15% vs indexflat · 27 dcaptured Deploy ↗
Geodd$3.00−15% vs indexflat · 22 dcaptured Deploy ↗
Gcore$24.45/node ÷ 8 · Bare Metal GPU Clusters · €2.73 listed · ECB 1.1205 on October 22, 2026$3.06−13% vs index+20.2% · 26 dcaptured Deploy ↗
Massed Compute$3.14−11% vs index+8.7%captured Deploy ↗
Hyperbolicfrom$3.19−10% vs indexflat · 26 dcaptured Deploy ↗
Hyperstack$3.20−9% vs indexflatcaptured Deploy ↗
Thunder Compute$3.20−9% vs indexflat · 24 dcaptured Deploy ↗
Canopy Wave$3.25−8% vs indexflat · 27 dcaptured Deploy ↗
Jarvislabs$3.49−1% vs index+29.7%captured Deploy ↗
Scaleway$28.35/node ÷ 8 · fr-par-2 · €3.17 listed · ECB 1.1193 on October 23, 2026$3.54+0% vs indexflatcaptured Deploy ↗
Elice Cloud$3.57+1% vs indexflat · 11 dcaptured Deploy ↗
Barrack AI$3.59+2% vs indexflat · 10 dcaptured Deploy ↗
Crusoe7d −13.3%$3.90+10% vs indexflatcaptured Deploy ↗
GPU.aifrom$3.99+13% vs index+14.3% · 22 dcaptured Deploy ↗
Lambda$3.99+13% vs indexflatcaptured Deploy ↗
RunPodAPISecure Cloud · the API request$3.99+13% vs index+14.3%captured Deploy ↗
Together AI$3.99+13% vs indexflatcaptured Deploy ↗
Neysafrom$4.39+24% vs indexflat · 27 dcaptured Deploy ↗
PaperspaceGPU Droplets$4.41+25% vs indexflatcaptured Deploy ↗
Nebius$4.50+27% vs indexflatcaptured Deploy ↗
VerdaAPI7d +7.2%$4.51+28% vs index+32.2%captured Deploy ↗
PaperspacePaperspace$5.95+69% vs indexflatcaptured Deploy ↗
CoreWeave$49.24/node ÷ 8$6.16+74% vs indexflatcaptured Deploy ↗
Amazon Web ServicesAPI$55.04/node ÷ 8 · us-east-1$6.88+95% vs indexflatcaptured Deploy ↗
STACKITGermany South · €8.04 listed · ECB 1.1193 on October 23, 2026$9.00+155% vs indexflat · 13 dcaptured Deploy ↗
Google CloudGPU onlyAPIus-west8 · cheapest of 10 regions read; dearest $11.78$9.80+178% vs indexflatcaptured Deploy ↗
Oracle Cloud (OCI)API$10.00+183% vs indexflatcaptured Deploy ↗
Microsoft AzureAPI$88.49/node ÷ 8 · eastus · cheapest of 24 regions read; dearest $23.33$11.06+213% vs indexflatcaptured Deploy ↗
Spot · high risk — 7 rates (1 not confirmed, out of every figure), cheapest first · Price Index $3.53
Provider · what is sold$ / GPU-hour30 daysReadSource
Nebiusfrom$0.79−78% vs indexunconfirmedlast confirmed · not in the figuresDeploy ↗
Google CloudGPU onlyAPIus-west2 · cheapest of 10 regions read; dearest $5.31$1.06−70% vs indexflatcaptured Deploy ↗
Microsoft AzureAPI$16.35/node ÷ 8 · eastus · cheapest of 24 regions read; dearest $4.31 · −$9.02 (−82%) vs its on-demand $11.06$2.04−42% vs indexflatcaptured Deploy ↗
VerdaAPI$4.51/node ÷ 2 · −$2.26 (−50%) vs its on-demand $4.51 · 7d +7.2%$2.26−36% vs index+32.2%captured Deploy ↗
CoreWeave$19.51/node ÷ 8 · Europe · −$3.72 (−60%) vs its on-demand $6.16$2.44−31% vs indexflatcaptured Deploy ↗
CoreWeave$19.71/node ÷ 8 · North America · −$3.69 (−60%) vs its on-demand $6.16$2.46−30% vs indexflatcaptured Deploy ↗
Elice Cloud−$0.36 (−10%) vs its on-demand $3.57$3.21−9% vs indexflat · 11 dcaptured Deploy ↗
Community · high risk — 6 rates (3 a thin market, out of every figure) (1 not confirmed, out of every figure), cheapest first · Price Index $3.53
Provider · what is sold$ / GPU-hour30 daysReadSource
LiumfloorAPI$1.39−61% vs indexthin marketfewer offers than a floor needs · last floor · not in the figuresDeploy ↗
Vast.aifloorAPI$3.58/node ÷ 2 · Multi-GPU machines$1.79−49% vs indexthin marketfewer offers than a floor needs · last floor · not in the figuresDeploy ↗
RunPodAPICommunity Cloud · the API request$2.69−24% vs indexflatcaptured Deploy ↗
GPU.aifrom$3.72+5% vs indexstoppedlast confirmed · not in the figures
LiumfloorAPI$30.00/node ÷ 8 · Multi-GPU machines$3.75+6% vs index+79.4% · 6 dcaptured Deploy ↗
Vast.aifloorAPI$4.07+15% vs indexthin marketfewer offers than a floor needs · last floor · not in the figuresDeploy ↗
Serverless · medium risk — 1 rate, cheapest first · Price Index $3.53
Provider · what is sold$ / GPU-hour30 daysReadSource
ModalGPU only$3.95+12% vs indexflatcaptured Deploy ↗
Reserved · low risk — 2 rates, cheapest first · Price Index $3.53
Provider · what is sold$ / GPU-hour30 daysReadSource
Hyperstackfrommin. commitment commit$2.72−23% vs indexflatcaptured Reserve ↗
Together AIfrommin. commitment commit$3.19−10% vs indexflatcaptured Reserve ↗
1-Year · low risk — 6 rates, cheapest first · Price Index $3.53
Provider · what is sold$ / GPU-hour30 daysReadSource
Seeweb€1.61 listed · ECB 1.1193 on October 23, 2026 · −$0.31 (−15%) vs its on-demand $2.12$1.80−49% vs indexflatcaptured Reserve ↗
Bitdeer AIfrom$4.15/node ÷ 2 · −$0.28 (−12%) vs its on-demand $2.36$2.08−41% vs indexflat · 13 dcaptured Reserve ↗
Civo−$0.30 (−10%) vs its on-demand $2.99$2.69−24% vs indexflat · 27 dcaptured Reserve ↗
PaperspaceGPU Droplets · −$1.15 (−26%) vs its on-demand $4.41$3.26−8% vs indexflatcaptured Reserve ↗
Google CloudGPU onlyAPIus-central1 · cheapest of 10 regions read; dearest $8.17$6.80+93% vs indexflatcaptured Reserve ↗
Microsoft AzureAPI$56.63/node ÷ 8 · eastus · cheapest of 24 regions read; dearest $14.93 · −$3.98 (−36%) vs its on-demand $11.06$7.08+101% vs indexflatcaptured Reserve ↗
3-Year · low risk — 4 rates, cheapest first · Price Index $3.53
Provider · what is sold$ / GPU-hour30 daysReadSource
Bitdeer AIfrom$3.59/node ÷ 2 · −$0.57 (−24%) vs its on-demand $2.36$1.79−49% vs indexflat · 13 dcaptured Reserve ↗
Civo−$0.50 (−17%) vs its on-demand $2.99$2.49−29% vs indexflat · 27 dcaptured Reserve ↗
Google CloudGPU onlyAPIus-east4 · cheapest of 10 regions read; dearest $5.17$4.30+22% vs indexflatcaptured Reserve ↗
Microsoft AzureAPI$38.85/node ÷ 8 · eastus · cheapest of 24 regions read; dearest $10.24 · −$6.21 (−56%) vs its on-demand $11.06$4.86+38% vs indexflatcaptured Reserve ↗

Other cards today, median rate card per GPU-hour: B200 $7.50 · H200 $4.56 · B300 $8.99 · GB300 NVL72 $18.00 · MI325X $2.25 · RTX 4090 $0.49 · all cards

Add your GPU cloudPrices come from your public pricing page, never from this form. Listing is free and can’t be bought. How providers are listed, and who is not, and why
Your email is used only about your listing, never shared.

Renting the H100 SXM for a year or more?

The cheapest committed H100 SXM rate each provider publishes, by term. A commitment is paid whether the GPU runs or not.

TermPer GPU-hourProvidervs its on-demand
1-year$1.80
€1.61 listed · ECB 1.1193 on October 23, 2026
read
Seeweb
cheapest of 6 providers
−15% ($2.12)
3-year$1.79
read
Bitdeer AI
cheapest of 4 providers
−24% ($2.36)
Other commitment
commit
$2.72
read
Hyperstack
cheapest of 2 providers
—

Across every card, a 1-year rate runs a median −15% against the same provider’s on-demand rate (7 providers); a 3-year rate −30% (5 providers). Every committed rate, card by card.

Get quotes for the H100 SXM. A person at GPU Economy reads your request and passes it to up to three providers that list a rate for the card, the cheapest listed first; the quotes come from them, directly to you.

Term

We may be paid a fee if a request becomes a contract; who receives it follows the listed rates, never a fee, and it changes no published rate or the Price Index (how, privacy).

As of , the cheapest listed rate for the NVIDIA H100 SXM is $1.06 per GPU-hour on Google Cloud (spot, us-west2). The cheapest on-demand rate is $1.79 on Cudo Compute (no-kristiansand-1, GPU only). Across 38 providers, the median rate card — each provider’s cheapest on-demand rate, community capacity and marketplace floors out — is $3.52 per GPU-hour, and the middle half of those rate cards sits between $2.99 and $4.29. Between rate cards, on-demand rates for the same card run from $1.79 (Cudo Compute) to $11.06 (Microsoft Azure), 6.2× apart across 38 providers. By market, the cheapest listed rates are $1.06 spot (Google Cloud, GPU only), $2.69 on community capacity (RunPod), $1.79 on-demand (Cudo Compute, GPU only), $2.72 reserved with no stated term (Hyperstack, from), $1.80 on a 1-year commitment (Seeweb) and $1.79 on a 3-year commitment (Bitdeer AI, from). On a single rate card, Microsoft Azure (eastus) lists it at $11.06 on-demand, $7.08 on a 1-year commitment, $4.86 on a 3-year commitment and $2.04 spot: a commitment is paid whether the GPU runs or not, and spot capacity can be reclaimed mid-run. To buy, the median new-condition listing is $36,457.50 across 4 sellers. With 80 GB of VRAM, that is $0.022 per GB of memory per hour at the cheapest on-demand rate and $0.044 at the median.

The two cheapest on-demand rate cards for the NVIDIA H100 SXM, side by side on every card they both list: Cudo Compute vs Seeweb.

Dated pieces on the NVIDIA H100 SXM: — HPC-AI, Barrack, Trooper first read October 13, 2026: H100 SXM $2.20 · — FPT and Elice first read October 12, 2026: H100 SXM $2.99 and $3.57 · — Bitdeer first read October 10, 2026: GB200 $6.94, H100 SXM $2.36. Each is frozen as of its date.

As of , 5 providers price the NVIDIA H100 SXM on-demand as a whole 8-GPU node (HGX, as CoreWeave, Crusoe, Nebius and Paperspace name it in their own listings): from $24.45 per node-hour (Gcore, 8×H100 SXM, Bare Metal GPU Clusters, $3.06 per GPU) to $88.49 (Microsoft Azure, Standard_ND96is_H100_v5, eastus, $11.06 per GPU), 3.6× apart. 8 GPUs at the cheapest on-demand rate ($1.79 on Cudo Compute, GPU only) come to $14.32 an hour, where a provider rents 8 to one account and they need not share a unit. A unit price is what the provider quotes for the 8 GPUs together; whether it also rents one is not part of the quote. 1 prices it in units of 2 — the rate table below names each quote's own unit.

How it is computed

What we measure

Publicly listed rental prices for 79 GPU models (NVIDIA and AMD) at 99 providers — 1165 current quotes across seven markets — and the retail price of the cards.

Where it comes from

Each provider's own pricing page or API; retail from reseller pages. Every figure links back to its source.

How it is normalised

Node prices are divided by GPU count, per-second rates multiplied out, a price listed in another currency shown as listed, beside its USD figure at the dated official rate (ECB; the Bank of Russia for RUB and VND). A reservation published as a total for its term is divided by the hours in that term, 8,760 per year.

What is modelled, not measured

One figure: cost per million output tokens, a listed rate divided by the throughput NVIDIA published for that card, published with the run that produced it in /token-cost.json, not on this page. Risk is not a model — it is read off the market a quote is from.

What we never do

Never estimate a price, never convert a currency silently, never publish a number a reader cannot verify on the linked page.

1165 quotes · 99 providers. Seven files under CC BY 4.0; history is stored append-only and never rewritten.

Weekly history since 2025, with a like-for-like table per provider: NVIDIA H100 SXM price history.

Retail listings

Median new-condition store price: $36,457.50 across 4 listings from 4 sellers, most recent September 28, 2026.

Current retail listings, last 30 days: condition and kind of price (a store’s own price, or a marketplace seller’s asking price). Each price links to the listing.
SellerCondition, kindPriceSeenSource
tes-itsolutions.comnew store price$22,753.50Listing
SPW Industrialnew store price$32,960Listing
NETGeneticsnew store price$39,955Listing
Buy Direct & Savenew store price$40,248.85Listing

At the cheapest on-demand rate ($1.79/hour on Cudo Compute, no-kristiansand-1, GPU only), that purchase price equals about 20,367 GPU-hours of rental — roughly 28 months of continuous use. The lower spot rate ($1.06 on Google Cloud) is left out: that capacity can be reclaimed or moves with supply. Power, hosting, cooling and idle time are not counted.

Compare the NVIDIA H100 SXM with: NVIDIA H100 PCIe · NVIDIA H200 · NVIDIA A100 80GB · NVIDIA B200 · AMD Instinct MI300X

Every rate on this page as data: /gpu/h100-sxm/prices.json — the same fields as prices.json, narrowed to this card. CC BY 4.0, no key, no rate limit. On your own page: embed this card’s cheapest rates (/embed/gpu/h100-sxm, an iframe, no script).

Data updated: