NVIDIA A100 40GB — what a GPU-hour costs
Listed rates, not availability. Read . · Renting for a year or more? Reserved quotes ↓
In September 2026 the A100 40GB like-for-like median did not move. Over 12 months, +41%.
Size your workload
Who is cheapest today, and what a month costs.
GPUsHours per monthRegionA100 40GB by provider
One table per market, on-demand first; cheapest first within each.
| Provider · what is sold | $ / GPU-hour | 30 days | Read | Source |
|---|---|---|---|---|
 Trooper.AIfrom€0.61 listed · ECB 1.1193 on October 23, 2026 | $0.68−66% vs median | flat · 10 d | captured | Deploy ↗ |
 centronGPU only€0.68 listed · ECB 1.1193 on October 23, 2026 | $0.76−62% vs median | flat · 13 d | captured | Deploy ↗ |
 SynpixCloud | $0.79−60% vs median | flat · 23 d | captured | Deploy ↗ |
 Jarvislabs | $0.89−55% vs median | flat | captured | Deploy ↗ |
 Civo | $1.09−45% vs median | flat · 28 d | captured | Deploy ↗ |
 Denvr Dataworks | $1.15−42% vs median | flat | captured | Deploy ↗ |
 NorthflankGPU only | $1.42−28% vs median | flat | captured | Deploy ↗ |
 Sestercefrom | $1.54−22% vs median | flat · 15 d | captured | Deploy ↗ |
 VerdaAPI7d +6.1% | $1.56−21% vs median | +21.8% | captured | Deploy ↗ |
 E2E Networks | $1.98−0% vs median | flat | captured | Deploy ↗ |
 EmpirioLabs | $1.99+0% vs median | flat · 28 d | captured | Deploy ↗ |
 Lambda | $1.99+0% vs median | flat | captured | Deploy ↗ |
 Amazon Web ServicesAPI$21.96/node ÷ 8 · us-east-1 | $2.74+38% vs median | flat | captured | Deploy ↗ |
 Google CloudGPU onlyAPIus-central1 · cheapest of 10 regions read; dearest $3.46 | $2.93+48% vs median | flat | captured | Deploy ↗ |
 Oracle Cloud (OCI)API | $3.05+54% vs median | flat | captured | Deploy ↗ |
 PaperspacePaperspace | $3.09+56% vs median | flat | captured | Deploy ↗ |
 Cloud.ru Evolution256.20 RUB listed · Bank of Russia 0.012416 on October 24, 2026 | $3.18+60% vs median | flat · 25 d | captured | Deploy ↗ |
 Microsoft AzureAPI$27.20/node ÷ 8 · westus2 · cheapest of 6 regions read; dearest $4.42 | $3.40+71% vs median | flat | captured | Deploy ↗ |
| Provider · what is sold | $ / GPU-hour | 30 days | Read | Source |
|---|---|---|---|---|
| Google CloudGPU onlyAPIus-west2 · cheapest of 10 regions read; dearest $1.86 | $0.31−84% vs median | flat | captured | Deploy ↗ |
| Microsoft AzureAPI$5.98/node ÷ 8 · westus2 · cheapest of 6 regions read; dearest $0.97 · −$2.65 (−78%) vs its on-demand $3.40 | $0.75−62% vs median | flat | captured | Deploy ↗ |
| VerdaAPI−$0.78 (−50%) vs its on-demand $1.56 · 7d +6.1% | $0.78−61% vs median | +21.8% | captured | Deploy ↗ |
| Provider · what is sold | $ / GPU-hour | 30 days | Read | Source |
|---|---|---|---|---|
 Vast.aifloorAPI7d −18.0% | $0.42−79% vs median | −31.1% | captured | Deploy ↗ |
| Vast.aifloorAPI$1.26/node ÷ 2 · Multi-GPU machines | $0.63−68% vs median | thin marketfewer offers than a floor needs · last floor · not in the figures | Deploy ↗ | |
 GPU.aifrom | $0.67−66% vs median | stoppedlast confirmed · not in the figures | ||
 Clore.aifloorAPI$1.66/node ÷ 2 · Multi-GPU machines | $0.83−58% vs median | +216.7% · 17 d | captured | Deploy ↗ |
 RunPodAPICommunity Cloud · the API request | $1.00−50% vs median | flat | captured | Deploy ↗ |
| Provider · what is sold | $ / GPU-hour | 30 days | Read | Source |
|---|---|---|---|---|
 CerebriumGPU only | $2.00+1% vs median | flat | captured | Deploy ↗ |
 ModalGPU only | $2.10+6% vs median | flat | captured | Deploy ↗ |
| Provider · what is sold | $ / GPU-hour | 30 days | Read | Source |
|---|---|---|---|---|
| Civo−$0.20 (−18%) vs its on-demand $1.09 | $0.89−55% vs median | flat · 28 d | captured | Reserve ↗ |
| Google CloudGPU onlyAPIus-central1 · cheapest of 10 regions read; dearest $2.18 · −$1.09 (−37%) vs its on-demand $2.93 | $1.85−7% vs median | flat | captured | Reserve ↗ |
| Microsoft AzureAPI$18.83/node ÷ 8 · westus2 · cheapest of 6 regions read; dearest $3.06 · −$1.05 (−31%) vs its on-demand $3.40 | $2.35+19% vs median | flat | captured | Reserve ↗ |
| Provider · what is sold | $ / GPU-hour | 30 days | Read | Source |
|---|---|---|---|---|
| Civo−$0.40 (−37%) vs its on-demand $1.09 | $0.69−65% vs median | flat · 28 d | captured | Reserve ↗ |
| Google CloudGPU onlyAPIus-west8 · cheapest of 10 regions read; dearest $1.21 | $1.03−48% vs median | flat | captured | Reserve ↗ |
| Microsoft AzureAPI$10.88/node ÷ 8 · westus2 · cheapest of 6 regions read; dearest $1.77 · −$2.04 (−60%) vs its on-demand $3.40 | $1.36−31% vs median | flat | captured | Reserve ↗ |
Other cards today, median rate card per GPU-hour: H100 SXM $3.53 (Price Index) · B200 $7.50 · H200 $4.54 · B300 $8.99 · GB300 NVL72 $18.00 · MI325X $2.25 · RTX 4090 $0.49 · all cards
Renting the A100 40GB for a year or more?
The cheapest committed A100 40GB rate each provider publishes, by term. A commitment is paid whether the GPU runs or not.
| Term | Per GPU-hour | Provider | vs its on-demand |
|---|---|---|---|
| 1-year | $0.89 read | Civo cheapest of 3 providers | −18% ($1.09) |
| 3-year | $0.69 read | Civo cheapest of 3 providers | −37% ($1.09) |
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 A100 40GB. 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.
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 A100 40GB is $0.31 per GPU-hour on Google Cloud (spot, us-west2). The cheapest on-demand rate is $0.68 on Trooper.AI (from). Across 20 providers, the median rate card — each provider’s cheapest on-demand rate, community capacity and marketplace floors out — is $1.98 per GPU-hour, and the middle half of those rate cards sits between $1.14 and $2.79. Between rate cards, on-demand rates for the same card run from $0.68 (Trooper.AI) to $3.40 (Microsoft Azure), 5.0× apart across 20 providers. By market, the cheapest listed rates are $0.31 spot (Google Cloud, GPU only), $0.42 on community capacity (Vast.ai, floor), $0.68 on-demand (Trooper.AI, from), $0.89 on a 1-year commitment (Civo) and $0.69 on a 3-year commitment (Civo). On a single rate card, Microsoft Azure (westus2) lists it at $3.40 on-demand, $2.35 on a 1-year commitment, $1.36 on a 3-year commitment and $0.75 spot: a commitment is paid whether the GPU runs or not, and spot capacity can be reclaimed mid-run. With 40 GB of VRAM, that is $0.017 per GB of memory per hour at the cheapest on-demand rate and $0.050 at the median.
The two cheapest on-demand rate cards for the NVIDIA A100 40GB, side by side on every card they both list: Trooper.AI vs centron.
As of , 2 providers price the NVIDIA A100 40GB on-demand as a whole 8-GPU node: from $21.96 per node-hour (Amazon Web Services, p4d.24xlarge, us-east-1, $2.74 per GPU) to $27.20 (Microsoft Azure, Standard_ND96asr_A100_v4, westus2, $3.40 per GPU), 1.2× apart. 8 GPUs at the cheapest on-demand rate ($0.68 on Trooper.AI, from) come to $5.46 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.
How it is computed
What we measure
Publicly listed rental prices for 79 GPU models (NVIDIA and AMD) at 99 providers — 1167 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.
1167 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 A100 40GB price history.
Retail listings
No retail listings tracked for this card in the last 30 days.
Compare the NVIDIA A100 40GB with: NVIDIA A100 80GB
Every rate on this page as data: /gpu/a100-40/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/a100-40, an iframe, no script).
Data updated: