How much does it cost to rent an L40S per hour?

Between $1.09 and $3.51 per GPU-hour on the published on-demand prices of 13 companies, read on their own price pages 12 Oct 2026 – 13 Oct 2026. The middle price is $1.67: about half the companies charge less, about half more. RunPod was the cheapest and Replicate the dearest — 3.2× apart for the same chip.

What is the cheapest place to rent an L40S?

On fixed on-demand prices, RunPod at $1.09 per GPU-hour, read 13 Oct 2026. Spot, community and marketplace listings can go lower, but they can be interrupted or depend on someone else's machine, so they are listed separately and never ranked. Reserved and committed contracts are lower again — and nobody publishes what large buyers actually pay.

How much does an L40S cost per month?

Running one L40S around the clock (730 hours) costs $796 at the cheapest published on-demand price, $1,220 at the middle price and $2,562 at the dearest. A machine of eight costs eight times that. Storage, network and data transfer are usually extra and are not included.

Why is the same GPU so much cheaper at one company than another?

Because you are not only buying the chip. The cheapest end is often brokered or resold capacity, priced per GPU with everything else extra. The dearest end is a hyperscaler list price for a whole 8-GPU machine with its CPUs, memory and storage, which exists mainly to make the committed rate look like a discount. Between them sit differences in support, networking, region, billing and contract terms — and in who the counterparty is.

NVIDIA L40S 48GB · prices read 12 Oct 2026 – 13 Oct 2026

What does it cost to rent an L40S?

13 companies publish an on-demand price for the L40S. Read on their own price pages, the cheapest is $1.09 per GPU-hour at RunPod and the dearest $3.51 at Replicate — 3.2× apart for the same chip. The middle price is $1.67. Every price below links to the page it was read on.

Cheapest on-demand · per GPU-hour

$1.09

RunPod

Middle price · per GPU-hour

$1.67

of 13 published prices

Dearest on-demand · per GPU-hour

$3.51

Replicate · 3.2× the cheapest

What a month costs

The rental alone, at the cheapest, middle and dearest published on-demand price. Storage, network and data transfer are usually billed on top and are not included.

How much	Cheapest	Middle	Dearest
1 GPU for 100 hours	$109	$167	$351
1 GPU for a working month (160 h)	$174	$267	$562
1 GPU always on (730 h)	$796	$1,220	$2,562
8 GPUs always on (730 h)	$6,366	$9,759	$20,498

Work out your own L40S bill →Any number of hours, 1 or 8 GPUs, every company side by side.

Every published on-demand L40S price, cheapest first

One row per company: its own fixed on-demand price per GPU per hour, rentable on the day it was read. Ordered by price and nothing else. Each row says how its number was made and how fresh it is — current for a week, aging to 30 days, stale after that; the site will not build with an on-demand price older than 45 days.

#	Company	Tier	Per GPU-hour	Freshness	Source
1	RunPodL40S 48GB How this number was made none — page prices per GPU Secure Cloud Pod — RunPod's own on-demand list price per GPU-hour, billed per second; ranked from 13 Oct on Abdul's call (one rule for every company: a published hourly price is ranked). Re-read 13 Oct 2026, ~22:35 EDT (14 Oct ~02:35 UTC), on runpod.io/pricing with the Secure Cloud toggle (the default); unchanged since 9 Oct. Region: not stated (30+ regions)	On-demand	$1.09	Aging · 13 Oct 2026	their page ↗
2	Civo How this number was made 1-GPU instance (2x/4x/8x are exact multiples) Per hour, 1 x L40S 48GB Region: not stated on page	On-demand	$1.29	Aging · 12 Oct 2026	their page ↗
3	Crusoe How this number was made Per GPU-hour (L40S 48GB) Region: not stated on page	On-demand	$1.50	Aging · 12 Oct 2026	their page ↗
4	Nebius How this number was made 'from $1.55' per GPU-hour, L40S with Intel CPU (8-40 vCPUs); L40S with AMD CPU is 'from $1.82'. Same in both columns Region: not stated on page	On-demand	$1.55	Aging · 12 Oct 2026	their page ↗
5	Paperspace (DigitalOcean) How this number was made Already per GPU Per GPU per hour, GPU Droplet (1 GPU); billed per second, 5-minute minimum Region: not stated on page	On-demand	$1.57	Aging · 12 Oct 2026	their page ↗
6	DataCrunch (now Verda) How this number was made Already per GPU (1x instance) Per hour, 1x L40S 48GB Region: not stated on page	On-demand	$1.66	Aging · 12 Oct 2026	their page ↗
7	Vultr How this number was made Already per GPU 'Price starting at' $1.671/GPU/hr, same for 1, 2, 4 and 8 GPUs Region: not stated (FAQ: some locations priced differently; per-location prices are in the Control Panel)	On-demand	$1.67	Aging · 12 Oct 2026	their page ↗
8	OVHcloudL40S 48GB (l40s-90: 1 GPU) How this number was made l40s-90 $1.80/hr, 1 GPU, no division. USD, excl. taxes, hourly (monthly $1,296). Same price on the international page (GRA11). Region: US-EAST-VA-1 and US-WEST-OR-1 (OVHcloud US regions)	On-demand	$1.80	Aging · 12 Oct 2026	their page ↗
9	AWS (P5, P6, G6)L40S 48GB (g6e.xlarge: 1 GPU, 4 vCPU, 32 GiB) How this number was made g6e.xlarge on-demand Linux $1.861/hr, 1 GPU, no division. Read from the price file AWS's on-demand page loads (b0.p.awsstatic.com/pricing/2.0/meteredUnitMaps/ec2/USD/current/ec2-ondemand-without-sec-sel/US East (N. Virginia)/Linux/index.json, published 2026-10-08T18:48:50Z); the page's table is rendered from it. Billed per second. Smallest 1-GPU size; larger G6e sizes cost more per GPU because they bundle more vCPU/RAM (g6e.48xlarge $30.13118/hr ÷ 8 = $3.7664). Region: us-east-1 (US East, N. Virginia)	On-demand	$1.86	Aging · 12 Oct 2026	their page ↗
10	ModalNvidia L40SBase price: GPU time only, preemptible by default. Non-preemptible runs cost 3x; CPU and memory are extra. How this number was made Per GPU per second x 3600, no division. Per-second billing: $0.000542/sec x 3600. GPU only; CPU/memory extra. Base (preemptible) rate; non-preemptible = 3x. Region: Not pinned: base price lets Modal place the job; region selection costs 1.15-1.75x	On-demand	$1.95	Aging · 12 Oct 2026	their page ↗
11	CoreWeave How this number was made $18.00/hr for an 8-GPU L40S (48GB) instance ÷ 8 Instance price per hour Region: North America (Europe on-demand identical)	On-demand	$2.25	Aging · 12 Oct 2026	their page ↗
12	Oracle Cloud (OCI)L40S 48GB (BM.GPU.L40S.4 / BM.GPU.L40S-NC.4, bare metal, 4 GPUs) How this number was made Oracle prices per GPU (metric 'GPU Per Hour'), PAY_AS_YOU_GO $3.5, no division. BM.GPU.L40S.4, part B109479. The page's price cells are empty in its HTML and filled from Oracle's public price-list API (https://apexapps.oracle.com/pls/apex/cetools/api/v1/products/?partNumber=B109479&currencyCode=USD, lastUpdated 2026-10-08), which is where the figure was read. Pay-as-you-go. Region: Not region-specific (Oracle's USD price list carries one rate per part number)	On-demand	$3.50	Aging · 12 Oct 2026	their page ↗
13	ReplicateNvidia L40S (gpu-l40s) How this number was made 1-GPU hardware class, no division. Per-second billing: $0.000975/sec; Replicate prints $3.51/hr. Region: Not stated on the page	On-demand	$3.51	Aging · 12 Oct 2026	their page ↗

Listed apart, never ranked

Cheaper, but not a price you can count on (11)

Spot, preemptible, community and marketplace listings. They can be interrupted, or the price is set by an independent host and moves with them.

Vast.ai (L40S 48GB): $0.25–$0.53 per GPU-hour · Marketplace · Spot — can be interrupted · global marketplace · Aging · 9 Oct 2026 their page ↗
Vast.ai (L40S 48GB): $0.40–$0.53 per GPU-hour · Marketplace listing · global marketplace · Aging · 9 Oct 2026 their page ↗
Vast.ai (L40S 48GB): $0.40–$0.63 per GPU-hour · Marketplace on-demand (hosts vary) · global marketplace; cheapest host: Taiwan (unverified) · Aging · 9 Oct 2026 their page ↗
Nebius: $0.74 per GPU-hour · Spot — can be interrupted · Aging · 9 Oct 2026 their page ↗
RunPod (L40S 48GB): $0.79 per GPU-hour · Community cloud — other people's hardware · Aging · 9 Oct 2026 their page ↗
Prime Intellect (L40S 48GB PCIe): $0.82–$1.12 per GPU-hour · Marketplace on-demand (hosts vary) · US (low); EU west / US (high) · Aging · 9 Oct 2026 their page ↗
DataCrunch (now Verda): $0.83 per GPU-hour · Spot — can be interrupted · Aging · 12 Oct 2026 their page ↗
io.net (L40S 48GB): $0.84–$1.90 per GPU-hour · Marketplace on-demand (hosts vary) · US (low); France / Poland (high) · Aging · 9 Oct 2026 their page ↗
Spheron (L40S 48GB): $0.96 per GPU-hour · Marketplace on-demand (hosts vary) · Aging · 9 Oct 2026 their page ↗
CoreWeave: $0.99 per GPU-hour · Spot — can be interrupted · North America · Aging · 9 Oct 2026 their page ↗
AWS (P5, P6, G6) (L40S 48GB (g6e.xlarge, 1x)): $1.83 per GPU-hour · Spot — can be interrupted · us-east-1 (US East, N. Virginia) · Aging · 9 Oct 2026 their page ↗

Reserved, committed and promo prices (2)

A contract term or a limited offer, not a price for renting by the hour. Published ones are listed here; negotiated rates are lower and are not published by anyone.

Vultr: $0.85 per GPU-hour · Reserved — a contract term · Aging · 9 Oct 2026 their page ↗
Civo: $0.89–$1.19 per GPU-hour · Committed — a contract term · Aging · 9 Oct 2026 their page ↗

Listed, but not comparable

Scaleway (L40S 48GB (L40S-1-48G)) publishes 1.4699 EUR per GPU-hour only, read 9 Oct 2026. We do not convert a price at an exchange rate the company did not publish. their page ↗
Akash Network rents it without a public price — No provider listing on the Akash network at read time (absent from the gpu-prices feed).. their page ↗
CUDO Compute rents it without a public price — Not listed on CUDO's pricing page (which is quote-only throughout).. their page ↗
Genesis Cloud rents it without a public price — Company page unreachable on 2026-10-09 (TLS error / Cloudflare 522); no price read.. their page ↗
SF Compute rents it without a public price — SF Compute prices sit behind a sign-in (pricing page redirects to login).. their page ↗
TensorDock rents it without a public price — No public price; the live marketplace that would show it needs a login.. their page ↗

Questions

How much does it cost to rent an L40S per hour?: Between $1.09 and $3.51 per GPU-hour on the published on-demand prices of 13 companies, read on their own price pages 12 Oct 2026 – 13 Oct 2026. The middle price is $1.67: about half the companies charge less, about half more. RunPod was the cheapest and Replicate the dearest — 3.2× apart for the same chip.
What is the cheapest place to rent an L40S?: On fixed on-demand prices, RunPod at $1.09 per GPU-hour, read 13 Oct 2026. Spot, community and marketplace listings can go lower, but they can be interrupted or depend on someone else's machine, so they are listed separately and never ranked. Reserved and committed contracts are lower again — and nobody publishes what large buyers actually pay.
How much does an L40S cost per month?: Running one L40S around the clock (730 hours) costs $796 at the cheapest published on-demand price, $1,220 at the middle price and $2,562 at the dearest. A machine of eight costs eight times that. Storage, network and data transfer are usually extra and are not included.
Why is the same GPU so much cheaper at one company than another?: Because you are not only buying the chip. The cheapest end is often brokered or resold capacity, priced per GPU with everything else extra. The dearest end is a hyperscaler list price for a whole 8-GPU machine with its CPUs, memory and storage, which exists mainly to make the committed rate look like a discount. Between them sit differences in support, networking, region, billing and contract terms — and in who the counterparty is.

How these numbers were made, and how to reuse them

Macrostack does not rent clusters and does not benchmark them. Every figure is a published price read on the company's own price page on the date in its row. Committed, reserved and negotiated rates are lower than anything shown as a price here, and the real ones are not published by anyone.

Each price read on the company's own public price page on the date shown. On-demand per GPU per hour in USD; multi-GPU machines divided by GPU count (see normalization); per-second billing converted to per hour. Not measured by Macrostack. Committed, reserved and negotiated rates are lower and are not published by anyone. The full table — every chip, every company, 45 L40S rows among them — is free to reuse under CC BY 4.0 with credit and a link.

Download CSV Download JSON

Cite as

Macrostack (2026). L40S rental prices, read 13 Oct 2026. CC BY 4.0. https://www.macrostack.net/compute/gpu-prices/l40s

Every chip side by side: the GPU cost calculator · Other chips: H100 SXM rental prices · H200 rental prices · B200 rental prices · A100 80GB rental prices · L4 rental prices · Who these companies are: Layer 2.5 — Compute.

Prices read 12 Oct 2026 – 13 Oct 2026.