Rent L40: price today

Prices verified 2026-11-04 22:13 UTC (18:13 ET) · 3 offers from 3 vendors. Per GPU-hour, among tracked providers only.

Some links are referral links; Lantern Works may earn credit. Affiliate status is not a ranking factor.

On-demand

Link: “Referral link” means Lantern Works may earn credit if you sign up through it; “Direct link” goes to the vendor with no referral. Neither affects order. Vendor quote: a short excerpt of the vendor's pricing page; open it for the full text and source. Availability is hidden because no vendor here publishes a stock signal yet.

Spot & community

Link: “Referral link” means Lantern Works may earn credit if you sign up through it; “Direct link” goes to the vendor with no referral. Neither affects order. Vendor quote: a short excerpt of the vendor's pricing page; open it for the full text and source. Availability: Where a status is shown, it is a snapshot from the time in the row, not a guarantee. Statuses from a vendor's live stock check can change within minutes; “vendor-labelled” means a fixed label on the vendor's page, not a capacity check; “—” or “unknown” means no signal, not sold out. Region lists refresh with each Scout cycle and may be blank between cycles.

TensorDock not shown: held until re-verified (vendor pricing page last updated 2024-07-24).

Read before comparing

Rankings are the cheapest among tracked providers on our check date, not across the whole market. Vendors we don't track may cost less.

GPU-hours vs wall-clock time

Prices are per GPU-hour: one GPU for one hour. Four GPUs for 10 wall-clock hours is 40 GPU-hours. Per-second billing (serverless) charges only while your code runs; hourly pods bill from start to stop, idle time included; monthly servers bill a flat price whatever you use. A faster GPU can finish the same job in fewer GPU-hours, so the lowest hourly rate isn't always the lowest cost per job.

Hardware differences

The same GPU name can mean different hardware. SXM modules link GPUs over NVLink and run at higher power limits than PCIe cards, so an H100 SXM is usually faster than an H100 PCIe, especially for multi-GPU training. Interconnect between GPUs and nodes (NVLink, InfiniBand, plain Ethernet) is rarely listed on pricing pages; treat it as unknown unless the vendor states it. VRAM decides which models fit at all. Hardware as listed: L40 48 GB (not stated).

Egress and storage extras

Totals on this page exclude these. Each cell cites the vendor page it came from; "Not published" means the vendor doesn't publish it there.

Interruptible vs on-demand

Interruptible (preemptible or spot) capacity can be stopped by the vendor mid-job and is usually cheaper; on-demand runs until you stop it. Every rate listed here is on-demand; none is marked interruptible by its vendor.

FAQ

What is the cheapest L40 among tracked providers per hour?

Among tracked providers, Thunder Compute at $0.79 per GPU-hour on-demand (On-demand), verified 2026-11-04.

Is spot cheaper for L40?

The cheapest spot/community rate among tracked providers is $0.33 (Lium) vs $0.79 on-demand. Spot or interruptible capacity may be preempted.

Are egress and storage included?

No. Totals exclude storage, egress, extra vCPU and taxes; each vendor's published terms (or 'not published') are shown in the table.

How often are prices checked?

A daily job re-reads every vendor pricing page; a rate not re-verified for 7 days is held and never ranked.

More

Other GPUs: A100 · B200 · B300 · H100 · H200 · L4 · L40S · MI300X · RTX 4090 · RTX 5090 · RTX 6000 Ada · RTX A4000 · RTX A6000 · RTX PRO 6000 Server · V100

GPU price hub · price changes · Data source: gpu-prices.json · gpu-price-changes.json (RSS) · history.json