01AGPU Rental Prices· manual survey 2026-10-20 · automated rows 2026-10-20

L40S rental prices

Compare L40S provider listings, billing terms and dated evidence. Posted quotes and market-implied estimates are shown separately.

All GPU prices

L40S · reference price · Typical — 1 specialist cloud

$1.55/GPU·hr

n=1 · Nebiussingle-venue LIST price · specialist class · posted list/ask prices, not transactions

Cite

Copy-ready citation

L40S reference price — Typical — 1 specialist cloud: $1.55/GPU·hr (n=1 · Nebius), observed Oct 20, 2026 (posted list/ask prices, not transactions). Cite as: AI Compute Tracker, ACT.L40S.SPEC-MED@primary-on-demand-v1, https://aicomputetracker.com/gpu-rental-prices/l40s/, AI Compute Tracker (aicomputetracker.com), CC BY 4.0; third-party terms apply to data from GPU cloud pricing pages, accessed <date you visit>

Replace “<date you visit>” with your access date.

0.0%

vs Oct 19, 2026

Typical — 1 specialist cloud · n=1 · Nebius: $1.55/GPU·hr
Big-cloud list price · n=2 · AWS, Oracle OCI: $1.86/GPU·hr

L40S · current published quotes

Latest observations 2026-10-20; some quotes retained from 2026-10-14.

4 venues

Search providersSort by

4 of 4 quotes

Filters

Billing model

Table options

Your filters, sorting and columns are remembered on this device.

Current L40S prices per GPU-hour by provider and billing model
Provider	$ / GPU·hr	Billing	Observed	Details
Nebius — open source page	$1.55	On-demandCPU / memory affects total	2026-10-20	Details Provider type Specialist Pricing notes Advertised on-demand starting rate; CPU/RAM configuration affects total price. Source Source page · Nebius published pricing ↗ Configuration L40S with Intel CPU; 8–40 vCPUs and 32–160GB RAM range Region / billing Provider regions vary · On-demand
AWS (g6e) — open source page	$1.86	On-demand	2026-10-20	Details Provider type Hyperscaler Pricing notes The 1-GPU g6e.xlarge; larger g6e sizes cost more per GPU. Source Source page · AWS Price List API ↗ Configuration Per GPU within g6e.xlarge (1× L40S) Region / billing us-east-1 (US East, N. Virginia) · On-demand
Modal — open source page	$1.95	Serverless / per-secondInterruptible · GPU only · CPU / memory extra · Unpinned region	2026-10-20	Details Provider type Specialist Pricing notes GPU-only base rate $0.000542/second × 3,600, for preemptible execution with no region pinned; non-preemptible execution costs 3× and region selection 1.15–1.75× the base price; CPU, memory, and other resources billed separately. Source Source page · Modal published pricing ↗ Configuration L40S GPU task; one GPU; billed per second Region / billing Provider regions vary · Serverless / per-second
Oracle OCI — open source page	$3.50	On-demand4-GPU instance≈ $14.00 / instance·hr	2026-10-14	Details Provider type Hyperscaler Pricing notes Published USD per GPU-hour; multiply by four for the full bare-metal instance. Source Source page · Oracle OCI published pricing ↗ Configuration BM.GPU.L40S.4; 4 × L40S 48GB Region / billing OCI public USD price list; regional availability varies · On-demand

Dates, billing and evidence

Rates are per accelerator-hour and retain each venue’s billing model. Instance totals multiply the recorded accelerator count by the quoted rate; rounding may apply. They describe the recorded configuration, not a provider-wide minimum. Unavailable sources retain their last valid quote and original observation date. Open a row for its full configuration and source. All rows: Provisional — Dated source-page observation; the raw payload is not yet pinned to the archive.

L40S · Marketplace floor

—

No marketplace quote

L40S · market coverage

4 venues

All-market access $1.55 → $3.50

L40S · Big-cloud list price

$1.86/GPU·hr

AWS (g6e) · n=2

Comparable price history

Loading evidence ledger…

1W · 1M · 3M · 1Y · ALL

L40S · daily venue-class bands

6 archived observation days — “Big-cloud list price”, “Typical — median of specialist clouds” and “Marketplace floor” per GPU-hour, each with its venue count (n)

Data

/price-history.json

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.json

/price-history/l40s.json

Per-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.json

No 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”. Third-party source data — such as Kalshi market data and providers’ list prices — remains subject to its sources’ terms.

Live quotes2026-10-20

Full data contract →

Cite

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 2026-10-20, https://aicomputetracker.com/gpu-rental-prices/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, accessed <date you visit>

Replace “<date you visit>” with your access date.

Embed

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.

Preview the embed →

Source and methodology

Daily class aggregates from published Neon quote snapshots, archived through 2026-10-20; 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 · 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: $4.28

$4.28 per dense BF16 PFLOP-hour: Nebius's $1.55/GPU·hr — the lowest specialist on-demand list price, observed 2026-10-20 — ÷ 0.36205 PFLOPS dense BF16.

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 20, 2026 survey: the big-cloud list price is $1.86/GPU·hr; the typical price is $1.55/GPU·hr, one specialist cloud's list price (Nebius). All figures are posted list/ask prices, not transactions.

Where is L40S cheapest to rent?

As of the Oct 20, 2026 survey, Nebius lists the cheapest on-demand price at $1.55/GPU·hr (specialist class); no marketplace floor is currently quoted. These are posted list/ask prices, not transactions.

Sources and methodology

Each provider row carries its observation date, billing terms and evidence. Historical charts keep their own sources, gaps, table view and CSV download.

Price methodology Evidence ledger This GPU’s history · JSON

←All GPUs & providers

Background reading: NVIDIA L40S on AI Wiki. AI Wiki is run by the same team as AI Compute Tracker.