Lambda review

Facts from Lambda’s own pricing and docs pages, September 202656 read from a quoted sentence on the page we archived.

Summary

Lambda rents GPUs from $0.69 per GPU-hour on demand, bills per minute, and asks no minimum commitment. Egress is free. Lambda does not publish an uptime SLA.

SOC 2 Type IIVolume discountsFree egressSSO
From
$0.69/GPU-hr
Billing
Per minute
Minimum commitment
None on demand
Egress
Free
Uptime SLA
not stated
Registered in
🇺🇸
Visit Lambda
On-demand pricing
×8, on demand · NVIDIA H100 SXM 80GB
$3.99/hr
Deploy
×8, on demand
$6.69/hr
Deploy
×8, on demand · NVIDIA A100 SXM 80GB
$2.79/hr
Deploy
×4, on demand · NVIDIA A6000 48GB
$1.09/hr
Deploy
×1, on demand
$3.29/hr
Deploy

See all 26 Lambda offers →

What’s good
Lowest published on-demand rate is $0.69/GPU-hr (NVIDIA Quadro RTX 6000 24GB), with no minimum commitment
Egress is free — no per-GB charge on data leaving
Billing is per minute
Self-serve signup — no sales call to get started
Commitment discounts: reserved capacity at the lowest prices
SOC 2 Type II
Plans
On-Demand instances1-Click Clusters
From$0.69/GPU-hr (Quadro RTX 6000)

Lambda · On-Demand instances

Every on-demand rate we hold for Lambda, per GPU per hour, cheapest first.

×8, on-demand
$0.79/hr
Deploy
×4, on-demand · NVIDIA A6000 48GB
$1.09/hr
Deploy
×1, on-demand
$1.29/hr
Deploy
×4, on-demand
$1.99/hr
Deploy
×8, on-demand · NVIDIA A100 SXM 40GB
$1.99/hr
Deploy
×1, on-demand
$2.29/hr
Deploy
×8, on-demand · NVIDIA A100 SXM 80GB
$2.79/hr
Deploy
×1, on-demand
$3.29/hr
Deploy
×8, on-demand · NVIDIA H100 SXM 80GB
$3.99/hr
Deploy
×8, on-demand
$6.69/hr
Deploy

Prices as last read from Lambda’s own pages. Open a card for every provider renting it.

$5.54–6.16/GPU-hr (H100 by size)

Lambda · 1-Click Clusters

Every on-demand rate we hold for Lambda, per GPU per hour, cheapest first.

×8, on-demand
$0.79/hr
Deploy
×4, on-demand · NVIDIA A6000 48GB
$1.09/hr
Deploy
×1, on-demand
$1.29/hr
Deploy
×4, on-demand
$1.99/hr
Deploy
×8, on-demand · NVIDIA A100 SXM 40GB
$1.99/hr
Deploy
×1, on-demand
$2.29/hr
Deploy
×8, on-demand · NVIDIA A100 SXM 80GB
$2.79/hr
Deploy
×1, on-demand
$3.29/hr
Deploy
×8, on-demand · NVIDIA H100 SXM 80GB
$3.99/hr
Deploy
×8, on-demand
$6.69/hr
Deploy

Prices as last read from Lambda’s own pages. Open a card for every provider renting it.

Minimum termNone2 weeks to 1 year; 1 year+ by quote
BillingPer minute, weekly invoicePer GPU-hour, billed weekly per the reservation
TenancyVMs, 1× to 8× GPUsSingle-tenant clusters, 16 to 2,000+ GPUs, InfiniBand
ProvisioningMinutes; first-come accessReservation reviewed and approved
PaymentCredit card ($10 pre-authorisation)Invoice, due within ten days
Details
Billing & access
How you pay
Postpaid, weekly invoices; clusters invoiced on reservation
Currency
USD, plus sales tax/VAT/GST
Commitment discounts
Reserved capacity at the lowest prices, by contact
Approval or KYC
Card pre-authorisation; purchases limited to listed countries
more
Australia, Canada, Chile, EU (except Lithuania), Iceland, Israel, Japan, New Zealand, Norway, Puerto Rico, Singapore, South Africa, South Korea, Switzerland, Taiwan, Turkey, UAE, UK, US.
Refunds
Service credits only; expire after 12 months
more
Credits cannot be redeemed for cash.
Late payment
Access may be suspended or terminated
Provisioning guarantee
First-come on demand; reserved clusters by approval
What you get
GPUs listed
B200, H100 SXM, H100 PCIe, GH200, A100 80GB, A100 40GB, A10, A6000, V100, Quadro RTX 6000
more
On-demand; 1-Click Clusters add HGX B200 and H100 at 16+ GPUs; GB300 NVL72 on Superclusters.
Regions
14 regions: US (9), Japan (2), India, Germany, Israel
more
asia-northeast-1 Tokyo, asia-northeast-2 Osaka, asia-south-1 India, europe-central-1 Germany, me-west-1 Israel, us-east-1/2, us-midwest-1, us-south-1/2/3, us-west-1/2/3.
Node configuration
8× H100 SXM: 208 vCPU · 1,800 GiB · 22 TiB SSD
Interconnect
SXM/NVLink within a node; InfiniBand on clusters
Local storage
Root volume per instance type (512 GiB – 22 TiB SSD)
Persistent storage
Filesystems, per GiB-month in hourly increments
Ingress
Free
Public endpoints
Firewall rulesets, global or per instance
Images
Ubuntu 22.04/24.04 with Lambda Stack (CUDA, PyTorch, TensorFlow, JAX); GPU Base; plain Ubuntu
Access methods
Console, SSH, JupyterLab, API, CLI
Model deployment
Self-managed on instances; observability built in
Managed Kubernetes
Managed Kubernetes
Kubernetes integration
Managed Kubernetes on clusters
FAQ
How much does Lambda cost?

As of September 2026, the lowest published on-demand rate is $0.69 per GPU-hour for the NVIDIA Quadro RTX 6000 24GB. Lambda publishes an on-demand rate for 11 GPU models. Billing is per minute.

Does Lambda have an uptime SLA?

Lambda does not publish an uptime commitment on its own site.

Is there a minimum commitment on Lambda?

No. Lambda states none on demand; 2 weeks on clusters.

Does Lambda charge for egress?

No. Lambda states that egress is free.

Can I sign up to Lambda without approval?

Lambda states that signup is self-serve. On approval and identity checks it states: card pre-authorisation; purchases limited to listed countries.

Are Lambda payments refundable?

Lambda states: service credits only; expire after 12 months.

Where are Lambda's data centres?

Lambda lists 14 regions: US (9), Japan (2), India, Germany, Israel.

Who do you contract with at Lambda?

The contracting entity is Lambda, Inc. (Delaware).

Reviewed September 2026, from Lambda’s own site. 56 facts, each quoted from the page it was read on.