CoreWeave · gpu
CoreWeave NVIDIA L40S pricing and specs
2 published prices for NVIDIA L40S in 2 regions across 1 billing term, from CoreWeave's official pricing source. Prices stay in native currency and can exclude storage, IPv4, egress, support and tax.
Price sources dated . Every price is within its freshness deadline. A site rebuild is not a new source check.
NVIDIA L40S specifications
| Family | gpu |
|---|---|
| Machine type | Virtual machine |
| vCPU | 128 |
| Physical cores | Unknown |
| Memory | 1024 GiB |
| Architecture | x86-64 |
| CPU allocation | Unknown |
| GPU | 8× NVIDIA L40S · 48 GiB/card |
| Local storage | Unknown |
| Network | Unknown |
| Provider status | Active |
Lowest and highest price by billing term
Each row compares one billing term in one currency across the regions that list it. Billing terms and currencies are never mixed or converted, and a stale price is compared only when no fresh one exists.
| Billing term | Lowest | Highest | Compared |
|---|---|---|---|
| On-demand | USD 18.00 / hrEurope price table (no data center stated) and 1 other at this price | Same as lowestSame price in all 2 regions compared | 2 regions |
Every published NVIDIA L40S price
One table per billing term, cheapest first. Monthly figures follow the provider's billing: a vendor monthly cap is not the hourly price multiplied by 730.
On-demand
2 prices
| Region | Hourly | Monthly | Tax and source date |
|---|---|---|---|
| Europe price table (no data center stated) | USD 18.00 / hr | ≈$13,140.00 / mo730-hour estimate, not a cap | Tax treatment not specifiedSource |
| North America price table (no data center stated) | USD 18.00 / hr | ≈$13,140.00 / mo730-hour estimate, not a cap | Tax treatment not specifiedSource |
Source notes for these prices
Every row: The page does not state whether prices include tax. Spot prices and the inference-platform single-GPU price (footnote 2: inference platform customers only) are not published here.
- Europe price table (no data center stated): CoreWeave lists $18.00 per instance per hour on demand in its "REGION: EUROPE" table. The monthly amount is a 730-hour estimate.
- North America price table (no data center stated): CoreWeave lists $18.00 per instance per hour on demand in its "REGION: NORTH AMERICA" table. The monthly amount is a 730-hour estimate.
Prices are published in US dollars; no currency conversion is involved.
Similar CoreWeave instance types
Up to 6 other CoreWeave instance types whose vCPU count and memory are both within a factor of two of this one, closest first, GPU or accelerator types only. Equal vCPU and RAM do not mean equal performance.
| Instance type | vCPU | Memory | CPU | Lowest on-demand |
|---|---|---|---|---|
| NVIDIA L40 | 128 | 1024 GiB | Unknown · x86-64 | USD 10.00 / hr |
| NVIDIA L40S | 128 | 1024 GiB | Unknown · x86-64 | USD 18.00 / hr |
| NVIDIA RTX PRO 6000 Blackwell Server Edition (High Memory) | 128 | 1024 GiB | Unknown · x86-64 | USD 20.00 / hr |
| NVIDIA A100 | 128 | 2048 GiB | Unknown · x86-64 | USD 21.60 / hr |
| NVIDIA HGX B200 | 128 | 2048 GiB | Unknown · x86-64 | USD 68.80 / hr |
| NVIDIA HGX H100 | 128 | 2048 GiB | Unknown · x86-64 | USD 49.24 / hr |
| NVIDIA GB200 NVL72 | 144 | 960 GiB | Unknown · arm64 | USD 42.00 / hr |
Sources and evidence
Official CoreWeave pricing page "On-demand GPU instances" tables, one per region (North America and Europe): on-demand prices per instance per hour, shown with "$" and recorded as USD. The page does not state a tax basis. Monthly amounts are 730-hour estimates. Products listed as "Contact sales" have no published price and are not shown; spot and inference-platform prices are out of scope.
View vendor pricing at CoreWeave ↗ · CoreWeave product page ↗
Data sources for these prices
The original collection sources, for auditing our data. They may contain raw API/JSON data or require authentication; they are not buying pages.
- https://www.coreweave.com/pricing · 2 prices · retrieved
Specification notes
- CoreWeave pricing page, "On-demand GPU instances": NVIDIA L40S, 8 GPUs, 48 GB VRAM per GPU, 128 vCPUs, 1,024 GB system RAM, 7.68 TB local storage.
- GPU and system memory are stored as the page's stated GB figures. The 7.68 TB local storage is not converted, because the page's binary-unit footnote is about CoreWeave Cloud storage and does not say which TB the instance column uses.
- CPU architecture x86-64: CoreWeave's instance docs (https://docs.coreweave.com/platform/instances/gpu/gd-8xl40s-i128.md, reviewed 2026-10-23) list the CPU model "Intel Sapphire Rapids 8462Y+ (2.80 GHz)" for 128 vCPUs and 1,024 GB RAM, and its Grace CPU guide (https://docs.coreweave.com/products/sunk/run_workloads/run-jobs-on-grace-cpu.md) says Grace instances report aarch64 and Intel or AMD instances report x86_64.
- Shared/dedicated allocation, physical cores, GPU interconnect and network throughput are not stated in the table.
Read the methodology before purchasing. All CoreWeave instance types and prices · Search the complete VM catalog