CoreWeave NVIDIA L40 vs RunPod L40 · Community Cloud.
machine vs machine · prices as of
The same L40 at two providers, so the board is identical and everything else is not. RunPod L40 · Community Cloud takes 2 of 4 rounds against CoreWeave NVIDIA L40. It is also the cheaper of the two by the hour, at $0.690/GPU-hr against $1.25/GPU-hr.
NVIDIA L40
$1.25/GPU-hr
$10/hr for the machine · $1,680 a week · $0.784 spot
8x L40 · 48 GB each · $0.0260/GB-hr
L40 · Community Cloud
$0.690/GPU-hr
$0.69/hr for the machine · $116 a week
1x L40 · 48 GB each · $0.0144/GB-hr · Marketplace
Price per GPU-hour (lower wins)
$1.25/GPU-hr
$0.690/GPU-hr
Interruptible rate (lower wins)
$0.784/GPU-hr
not published
Price per GB of VRAM (lower wins)
$0.0260/GB-hr
$0.0144/GB-hr
VRAM per GPU
48 GB
48 GB
GPUs per instance
8x L40
1x L40
vCPU
128
8
RAM
1024 GB
94 GB
Included storage
7.7 TB
none · volumes extra
Regions listed
2 regions
regions not stated
What a run costs on each
- An hour of the machine$10/hr vs $0.69/hr
- A day$240 vs $17
- A week$1,680 vs $116
- A week on spot$1,053 vs not published
CoreWeave then RunPod, at the published rate with no commitment discount
Go further
Add a third or fourth machine, or switch the weighting, in the tool.
Frequently asked
- Which is cheaper, CoreWeave NVIDIA L40 or RunPod L40 · Community Cloud?
- RunPod L40 · Community Cloud, at $0.690/GPU-hr against $1.25/GPU-hr, both the on-demand rate for one GPU for one hour, read on 17 Sept 2026. On the whole machine that is $0.69/hr against $10/hr.
- Which holds more GPU memory?
- CoreWeave NVIDIA L40, at 384 GB across the instance against 48 GB. Per board it is 48 GB against 48 GB.
- What would a week cost?
- $1,680 for CoreWeave NVIDIA L40 and $116 for RunPod L40 · Community Cloud, at 168 hours of the whole machine at its published rate, with no commitment discount. Where a spot rate is published the same week costs less and the machine can be taken back mid-run.
- What does the hourly price leave out?
- CoreWeave NVIDIA L40: disk 7.7 TB, 2 regions listed, billed per hour. RunPod L40 · Community Cloud: disk none, billed separately, no regions published, billed per second. Egress and tax are extra on both.
Every rate is the on-demand price for one GPU for one hour, read 17 Sept 2026 from each provider's own page, so a node of eight is compared at what one of its GPUs costs. Storage, egress and tax are extra unless a row says otherwise, and reserved or committed rates are not indexed. Bars are scaled to the larger value in each row.