NVIDIA L40S — what a GPU-hour costs
Listed rates, not availability. 38 rates read between and . · Renting for a year or more? Reserved quotes ↓
In September 2026 the L40S like-for-like median did not move. Over 12 months, 0%.
Size your workload
Who is cheapest today, and what a month costs.
GPUsHours per monthRegionL40S by provider
One table per market, on-demand first; cheapest first within each.
| Provider · what is sold | $ / GPU-hour | 30 days | Read | Source |
|---|---|---|---|---|
 CloudRift | $0.63−55% vs median | flat · 7 d | captured | Deploy ↗ |
 Beamfrom | $0.76−46% vs median | flat | captured | Deploy ↗ |
 GPU.aifrom | $0.80−43% vs median | flat · 22 d | captured | Deploy ↗ |
| Infomaniak0.70 CHF listed · ECB 1.213070 on October 23, 2026 | $0.85−39% vs median | flat · 14 d | captured | Deploy ↗ |
 Cudo ComputeGPU onlyAPIno-kristiansand-1 | $0.87−38% vs median | flat | captured | Deploy ↗ |
 packet.aiDedicated | $0.92−34% vs median | flat · 27 d | captured | Deploy ↗ |
 Seeweb€0.85 listed · ECB 1.1193 on October 23, 2026 | $0.95−32% vs median | flat | captured | Deploy ↗ |
 Hostingerfrom | $0.96−31% vs median | flat · 21 d | captured | Deploy ↗ |
 Edgevana$1.93/node ÷ 2 | $0.96−31% vs median | stoppedlast confirmed · not in the figures | ||
 Massed Compute | $0.97−30% vs median | flat | captured | Deploy ↗ |
 Sestercefrom7d +0.3% | $0.97−30% vs median | +27.6% | captured | Deploy ↗ |
 Geodd | $1.00−28% vs median | flat · 22 d | captured | Deploy ↗ |
 HexGridSecure Cloud | $1.00−28% vs median | flat · 27 d | captured | Deploy ↗ |
 QuantaCloudMidwest - us-midwest-2 | $1.09−22% vs median | stoppedlast confirmed · not in the figures | ||
| QuantaCloudMidwest - us-midwest-1 | $1.09−22% vs median | flat · 18 d | captured | Deploy ↗ |
 RunPodAPISecure Cloud · the API request | $1.09−22% vs median | flat | captured | Deploy ↗ |
 E2E Networks | $1.20−14% vs median | flat | captured | Deploy ↗ |
 HikubeGPU only1.04 CHF listed · ECB 1.213070 on October 23, 2026 | $1.26−9% vs median | flat · 15 d | captured | Deploy ↗ |
 Civo | $1.29−8% vs median | flat · 27 d | captured | Deploy ↗ |
 Gcore$11.41/node ÷ 8 · Bare Metal GPU Clusters · €1.27 listed · ECB 1.1205 on October 22, 2026 | $1.43+2% vs median | +16.7% · 26 d | captured | Deploy ↗ |
 TensorPoolGPU only | $1.49+7% vs median | stoppedlast confirmed · not in the figures | ||
 Crusoe | $1.50+8% vs median | flat | captured | Deploy ↗ |
 Nebiusfrom | $1.55+11% vs median | flat | captured | Deploy ↗ |
 Scalewayfr-par-2 · €1.47 listed · ECB 1.1193 on October 23, 2026 | $1.65+18% vs median | flat | captured | Deploy ↗ |
| Scalewaypl-waw-2 · €1.47 listed · ECB 1.1193 on October 23, 2026 | $1.65+18% vs median | flat | captured | Deploy ↗ |
 Atlantic.Net | $1.67+20% vs median | flat | captured | Deploy ↗ |
 EmpirioLabs | $1.80+29% vs median | flat · 27 d | captured | Deploy ↗ |
 Hugging Faceaws | $1.80+29% vs median | flat · 27 d | captured | Deploy ↗ |
 OVHcloudAPIUS | $1.80+29% vs median | flat | captured | Deploy ↗ |
 VESSL AI | $1.80+29% vs median | flat · 12 d | captured | Deploy ↗ |
 VerdaAPI$14.47/node ÷ 8 · 7d +5.1% | $1.81+30% vs median | +20.8% | captured | Deploy ↗ |
 Amazon Web ServicesAPIus-east-1 | $1.86+33% vs median | flat | captured | Deploy ↗ |
 Neysafrom | $1.95+40% vs median | flat · 27 d | captured | Deploy ↗ |
 STACKITGermany South · €1.74 listed · ECB 1.1193 on October 23, 2026 | $1.95+40% vs median | flat · 13 d | captured | Deploy ↗ |
 Lyceum | $1.99+43% vs median | +67.2% · 27 d | captured | Deploy ↗ |
 CoreWeave$18.00/node ÷ 8 | $2.25+61% vs median | flat | captured | Deploy ↗ |
 MimicPC7d +30.2% | $2.59+86% vs median | +30.2% · 9 d | captured | Deploy ↗ |
 Oracle Cloud (OCI)API | $3.50+151% vs median | flat | captured | Deploy ↗ |
 Replicate | $3.51+152% vs median | flat | captured | Deploy ↗ |
| Provider · what is sold | $ / GPU-hour | 30 days | Read | Source |
|---|---|---|---|---|
| Nebiusfrom | $0.74−47% vs median | unconfirmedlast confirmed · not in the figures | Deploy ↗ | |
| Lyceum | $0.79−43% vs median | stoppedlast confirmed · not in the figures | ||
| VerdaAPI−$0.90 (−50%) vs its on-demand $1.81 · 7d +5.1% | $0.90−35% vs median | +20.8% | captured | Deploy ↗ |
| CoreWeave$7.88/node ÷ 8 · North America · −$1.27 (−56%) vs its on-demand $2.25 | $0.98−29% vs median | flat | captured | Deploy ↗ |
 VultrAPI$11.92/node ÷ 8 | $1.49+7% vs median | flat · 24 d | captured | Deploy ↗ |
| MimicPC−$1.00 (−39%) vs its on-demand $2.59 · 7d +33.6% | $1.59+14% vs median | +33.6% · 9 d | captured | Deploy ↗ |
| Provider · what is sold | $ / GPU-hour | 30 days | Read | Source |
|---|---|---|---|---|
| GPU.aifrom | $0.79−43% vs median | flat · 22 d | captured | Deploy ↗ |
| RunPodAPICommunity Cloud · the API request | $0.79−43% vs median | flat | captured | Deploy ↗ |
 Vast.aifloorAPI | $0.80−43% vs median | thin marketfewer offers than a floor needs · last floor · not in the figures | Deploy ↗ | |
| Vast.aifloorAPI$3.34/node ÷ 4 · Multi-GPU machines · 7d +4.1% | $0.83−40% vs median | +4.1% | captured | Deploy ↗ |
| Provider · what is sold | $ / GPU-hour | 30 days | Read | Source |
|---|---|---|---|---|
 Koyeb | $1.20−14% vs median | flat | captured | Deploy ↗ |
| GcoreEverywhere Inference · €1.08 listed · ECB 1.1205 on October 22, 2026 | $1.21−13% vs median | flat · 26 d | captured | Deploy ↗ |
 CerebriumGPU only | $1.95+40% vs median | flat | captured | Deploy ↗ |
 ModalGPU only | $1.95+40% vs median | flat | captured | Deploy ↗ |
| Provider · what is sold | $ / GPU-hour | 30 days | Read | Source |
|---|---|---|---|---|
| CloudRiftmin. commitment 1mo · −$0.06 (−10%) vs its on-demand $0.63 | $0.57−59% vs median | flat | captured | Reserve ↗ |
| Provider · what is sold | $ / GPU-hour | 30 days | Read | Source |
|---|---|---|---|---|
| Seeweb€0.72 listed · ECB 1.1193 on October 23, 2026 · −$0.15 (−15%) vs its on-demand $0.95 | $0.81−42% vs median | flat | captured | Reserve ↗ |
| Civo−$0.20 (−16%) vs its on-demand $1.29 | $1.09−22% vs median | flat · 27 d | captured | Reserve ↗ |
| Atlantic.Net−$0.02 (−1%) vs its on-demand $1.67 | $1.65+18% vs median | flat | captured | Reserve ↗ |
| Provider · what is sold | $ / GPU-hour | 30 days | Read | Source |
|---|---|---|---|---|
| Civo−$0.40 (−31%) vs its on-demand $1.29 | $0.89−36% vs median | flat · 27 d | captured | Reserve ↗ |
| Atlantic.Net−$0.09 (−6%) vs its on-demand $1.67 | $1.58+13% vs median | flat | captured | Reserve ↗ |
Other cards today, median rate card per GPU-hour: H100 SXM $3.53 (Price Index) · B200 $7.50 · H200 $4.56 · B300 $8.99 · GB300 NVL72 $18.00 · MI325X $2.25 · RTX 4090 $0.49 · all cards
Renting the L40S for a year or more?
The cheapest committed L40S 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 | $0.81 €0.72 listed · ECB 1.1193 on October 23, 2026 read | Seeweb cheapest of 3 providers | −15% ($0.95) |
| 3-year | $0.89 read | Civo cheapest of 2 providers | −31% ($1.29) |
| Other commitment 1mo | $0.57 read | CloudRift | −10% ($0.63) |
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 L40S. 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 L40S is $0.57 per GPU-hour on CloudRift (reserved). The cheapest on-demand rate is $0.63 on CloudRift. Across 38 providers, the median rate card — each provider’s cheapest on-demand rate, community capacity and marketplace floors out — is $1.40 per GPU-hour, and the middle half of those rate cards sits between $0.98 and $1.85. Between rate cards, on-demand rates for the same card run from $0.63 (CloudRift) to $3.51 (Replicate), 5.6× apart across 38 providers. By market, the cheapest listed rates are $0.90 spot (Verda), $0.79 on community capacity (GPU.ai, from), $0.63 on-demand (CloudRift), $0.57 reserved with no stated term (CloudRift), $0.81 on a 1-year commitment (Seeweb) and $0.89 on a 3-year commitment (Civo). On a single rate card, Atlantic.Net lists it at $1.67 on-demand, $1.65 on a 1-year commitment and $1.58 on a 3-year commitment: a commitment is paid whether the GPU runs or not. To buy, the median new-condition listing is $8,999 across 9 sellers. With 48 GB of VRAM, that is $0.013 per GB of memory per hour at the cheapest on-demand rate and $0.029 at the median.
The two cheapest on-demand rate cards for the NVIDIA L40S, side by side on every card they both list: CloudRift vs Beam.
As of , 3 providers price the NVIDIA L40S on-demand as a whole 8-GPU node: from $11.41 per node-hour (Gcore, 8×L40S PCIe, Bare Metal GPU Clusters, $1.43 per GPU) to $18.00 (CoreWeave, L40S, $2.25 per GPU), 1.6× apart. 8 GPUs at the cheapest on-demand rate ($0.63 on CloudRift) come to $5.04 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 — 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 L40S price history.
Retail listings
Median new-condition store price: $8,999 across 13 listings from 9 sellers, most recent October 19, 2026.
| Seller | Condition, kind | Price | Seen | Source |
|---|---|---|---|---|
| ASA Computers | new store price | $7,569.10 | Listing | |
| ServerSupply.com | new store price | $8,260.25 | Listing | |
| ServerSupply.com | new store price | $8,545.25 | Listing | |
| Newegg.com | new store price | $8,595 | Listing | |
| ServerSupply.com | new store price | $8,787.50 | Listing | |
| Network Outlet | new store price | $8,900 | Listing | |
| newegg.com | new store price | $8,999 | Listing | |
| Newegg.com - xByte Technologies | refurbished marketplace listing | $8,999 | Listing | |
| Cloud Ninjas | new store price | $10,584 | Listing | |
| Orange Hardwares | new store price | $11,191.95 | Listing | |
| ServerOrbit.com | new store price | $11,485.50 | Listing | |
| DigiKey | new store price | $11,981.97 | Listing | |
| ServerSupply.com | new store price | $12,350 | Listing | |
| ServerSupply.com | new store price | $12,350 | Listing |
At the cheapest on-demand rate ($0.63/hour on CloudRift), that purchase price equals about 14,284 GPU-hours of rental — roughly 20 months of continuous use. The lower reserved rate ($0.57 on CloudRift) is left out: that capacity can be reclaimed or moves with supply. Power, hosting, cooling and idle time are not counted.
Compare the NVIDIA L40S with: NVIDIA RTX A6000 · NVIDIA RTX PRO 6000 Blackwell · NVIDIA RTX 4090
Every rate on this page as data: /gpu/l40s/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/l40s, an iframe, no script).
Data updated: