01AGPU cloud · accelerator ledger· manual survey 2026-10-14 · automated rows 2026-10-14
L40S rental prices
What capacity costs. Compare the active venue stack for L40S, the provider-class spread, and the long-run price direction where history is available.
L40S · reference price · Typical — median of 4 specialist clouds
$1.53/GPU·hr
Cite — copy a citation for the L40S reference price
Copy-ready citation
L40S reference price — Typical — median of 4 specialist clouds: $1.53/GPU·hr, observed between Oct 11 and Oct 14, 2026 (posted list/ask prices, not transactions). Cite as: AI Compute Tracker, ACT.L40S.SPEC-MED@primary-on-demand-v1, https://aicomputetracker.com/gpu-cloud/l40s/, AI Compute Tracker (aicomputetracker.com), CC BY 4.0; third-party terms apply to data from GPU cloud pricing pages, RunPod, accessed <date you visit>
Replace “<date you visit>” with your access date.
0.0%
vs Oct 13, 2026 · like-for-like
the 3 venues quoted in both; CoreWeave added
- Typical — median of 4 specialist clouds
- $1.53/GPU·hr
- Big-cloud list price · n=2 · AWS, Oracle OCI
- $1.86/GPU·hr
L40S · market coverage
7 venues
All-market access $1.09 → $3.50
L40S · Big-cloud list price
$1.86/GPU·hr
AWS (g6e) · n=2
L40S · specs and price per PFLOP-hour
Spec sources on /hardware →Spec-sheet row: L40S, reviewed 2026-10-15 against vendor documents; est marks a figure that review did not confirm. Dense tensor rates only, one precision per column — never a sparse figure.
- Dense BF16
- 362.05TFLOPS
- Dense FP8
- 733TFLOPS
- Dense FP4
- not listed on the spec sheet
- Memory
- 48GB
- Bandwidth
- 0.864TB/s
- TDP
- 350W
- $ / PFLOP·hr · BF16
- $3.01
$3.01 per dense BF16 PFLOP-hour: RunPod's $1.09/GPU·hr — the lowest specialist on-demand list price, observed — ÷ 0.36205 PFLOPS dense BF16.
Comparable price history
Loading evidence ledger…
L40S · daily venue-class bands
4 archived observation days — “Big-cloud list price”, “Typical — median of specialist clouds” and “Marketplace floor” per GPU-hour, each with its venue count (n)
More actions for L40S · daily venue-class bands
Data — endpoints, license, and freshness behind L40S · daily venue-class bands
The complete GPU price-history ledger: every series and dated observation, with billing, comparison, quality, and raw-evidence fields.
curl -s https://aicomputetracker.com/price-history.jsonPer-accelerator shard of the price-history ledger — the same series and point shapes filtered to one chip.
curl -s https://aicomputetracker.com/price-history/l40s.jsonNo API key, CORS open (Access-Control-Allow-Origin: *) — fetch it from a browser, a notebook, or a cron job.
The site’s compilation is licensed CC BY 4.0 — attribution for this chart’s data “AI Compute Tracker (aicomputetracker.com), CC BY 4.0; third-party terms apply to data from Amazon Web Services, Oracle Cloud Infrastructure, GPU cloud pricing pages, RunPod”. Third-party source data — such as Kalshi market data, Vast.ai marketplace data, and providers’ list prices — remains subject to its sources’ terms.
Live quotes
Cite — copy a citation for L40S · daily venue-class bands
Copy-ready citation
Cite as: AI Compute Tracker, “L40S · daily venue-class bands”, ACT.L40S.HYP-FLOOR@primary-on-demand-v1, ACT.L40S.SPEC-MED@primary-on-demand-v1 (posted list/ask prices, not transactions), live quotes as of 2026-10-14, https://aicomputetracker.com/gpu-cloud/l40s/#class-bands, AI Compute Tracker (aicomputetracker.com), CC BY 4.0; third-party terms apply to data from Amazon Web Services, Oracle Cloud Infrastructure, GPU cloud pricing pages, RunPod, accessed <date you visit>
Replace “<date you visit>” with your access date.
Embed — copy an iframe snippet for L40S · daily venue-class bands
Copy-paste iframe
The frame keeps this chart’s own freshness stamp, any third-party source credit, and the attribution line; the source line under it links back here with the data’s licence terms. It updates when the site redeploys.
Daily class aggregates from published Neon quote snapshots, archived through 2026-10-14; their citable series IDs, with each comparison-policy version, are under Cite. A day or class without a comparable quote stays a gap — never interpolated; a median on fewer than 3 venues is drawn as its range (a vertical bar), never as a point on the line, and the tooltip and table view print that range.
L40S · current published quotes
Hourly prices retain each venue’s billing model, cheapest first. Observation dates vary by row: unavailable sources retain their last valid quote, never a new observation date. All rows: Provisional — Dated source-page observation; the raw payload is not yet pinned to the archive.
| Provider | Class | Billing | $ / GPU·hr | Notes | Evidence | Observed |
|---|---|---|---|---|---|---|
| RunPod | Specialist | On-demand | $1.09 | Secure Cloud selected; community, serverless, and cluster prices differ. | Source — RunPod published pricing, RunPod | |
| Crusoe | Specialist | On-demand | $1.50 | GPU instance on-demand table; managed inference is a different product. | Source — Crusoe published pricing, Crusoe | |
| Nebius | Specialist | On-demand | $1.55 | Advertised on-demand starting rate; CPU/RAM configuration affects total price. | Source — Nebius published pricing, Nebius | |
| AWS (g6e) | Hyperscaler | On-demand | $1.86 | Published On-demand Linux rate for the 1-GPU g6e.xlarge; g6e.48xlarge (8 × L40S) is $30.13/hour, or $3.7664 per GPU. | Source — AWS (g6e) published pricing, AWS (g6e) | |
| Modal | Specialist | Serverless / per-second | $1.95 | GPU-only rate $0.000542/second × 3,600; CPU, memory, and other resources billed separately. | Source — Modal published pricing, Modal | |
| CoreWeave | Specialist | On-demand | $2.25 | Whole-instance on-demand price $18.00/hour divided by 8; single-GPU inference is a separate offering. | Source — CoreWeave published pricing, CoreWeave | |
| Oracle OCI | Hyperscaler | On-demand | $3.50 | Published USD per GPU-hour; multiply by four for the full bare-metal instance. | Source — Oracle OCI published pricing, Oracle OCI |
L40S rental price — common questions
Generated from the current datasets — each answer carries its own stamp and updates when the data refreshes.
How much does an L40S cost to rent?
As of the Oct 14, 2026 survey: the big-cloud list price is $1.86/GPU·hr; the typical price (median of 4 specialist clouds) is $1.53/GPU·hr (includes RunPod, last observed Oct 11, 2026). All figures are posted list/ask prices, not transactions.
Where is L40S cheapest to rent?
As of the Oct 14, 2026 survey, RunPod lists the cheapest on-demand price at $1.09/GPU·hr (specialist class, last observed Oct 11, 2026); no marketplace floor is currently quoted. These are posted list/ask prices, not transactions.
Background reading: NVIDIA L40S on AI Wiki. AI Wiki is run by the same team as AI Compute Tracker.