GPUBank vs Lambda
Lambda is a GPU cloud from San Jose, founded in 2012, that rents NVIDIA instances of 1 to 8 GPUs on demand and reserves InfiniBand clusters of 16 GPUs or more. Its H100 SXM costs $3.99 per GPU-hour on an 8-GPU instance and $4.29 on a single GPU; GPUBank charges $1.85. Lambda offers larger clusters and published SOC 2 and ISO certifications; GPUBank bills by the second, keeps a stopped instance at disk cost and takes crypto.
GPUBank
Tallinn, Estonia
H100 SXM
$1.85 / h
- vCPU, RAM, NVMe, IPv4 and traffic included
- Per-second billing after the health check
- Top up in Bitcoin, USDT and 12+ other coins
vCPU, RAM, NVMe, IPv4 and traffic included
Lambda
San Jose, United States · since 2012
H100 SXM
$3.99 / h
- Reserved InfiniBand clusters of 16 to 2,000+ GPUs
- B200 SXM6 on demand, GH200 at $2.29 per GPU-hour
- SOC 2 Type II and ISO 27001, 27017, 27701, 22301 stated
Lambda
GPU cloud
United States27 points compared23 of their pages readprices last checked on 29 September 2026page updated 2 November 2026
The verdict
Clusters and certifications at Lambda. Half the price and per-second billing at GPUBank.
Lambda is built for teams that need 8-GPU B200 or H100 nodes and multi-week InfiniBand clusters, with SOC 2 Type II and ISO 27001 paperwork. For the GPUs both of us rent, GPUBank costs 38 to 66 % less per GPU-hour. Lambda bills by the minute until an instance is terminated, since instances cannot be stopped; GPUBank bills by the second and keeps only the disk on the bill when you stop. Lambda takes major credit cards only; GPUBank takes crypto and asks for no card.
LambdaGPUBank 01The real cost
The same GPUs, side by side, per hour.
Every GPU model both of us rent, at each provider's own published price for one card, in US dollars, excluding VAT. Prices last checked on 29 September 2026.
| GPU | Lambda per hour | GPUBank per hour | Difference | 100 h at Lambda | 100 h at GPUBank |
|---|---|---|---|---|---|
 B200 SXM180 GB |
$6.69 | $4.15 | −38 % | $669.00 | $415.00 |
 H100 SXM80 GB |
$3.99 | $1.85 | −54 % | $399.00 | $185.00 |
 H100 PCIe80 GB |
$3.29 | $1.35 | −59 % | $329.00 | $135.00 |
 A100 SXM80 GB |
$2.79 | $0.95 | −66 % | $279.00 | $95.00 |
Lambda's lowest B200, H100 SXM and A100 SXM prices are per GPU on 8-GPU instances. A single H100 SXM costs $4.29 per hour there.
A fine-tune, 40 h on an H100 SXM
Lambda
$159.60
GPUBank
$74.00
Saves $85.60 at GPUBank (−54 %)
A training run, 100 h on an A100 SXM
Lambda
$279.00
GPUBank
$95.00
Saves $184.00 at GPUBank (−66 %)
A full day, 24 h on a B200
Lambda
$160.56
GPUBank
$99.60
Saves $60.96 at GPUBank (−38 %)
02Their catalogue
Every GPU Lambda lists.
Lambda lists 14 GPU models. The 13 below are the ones recorded in our check, each at its lowest public price per GPU-hour. 4 of them are also rented at GPUBank (in blue).
GB300reservedNo price- B300reservedNo price
- B200180 GB · on demand$6.69
GH20096 GB · on demand$2.29- H100 SXM80 GB · on demand$3.99
- H100 PCIe80 GB · on demand$3.29
- Vera Rubin NVL72reservedNo price
- A100 SXM 80GB80 GB · on demand$2.79
- A100 40GB40 GB · on demand$1.99
A1024 GB · on demand$1.29
V10016 GB · on demand$0.79
RTX A600048 GB · on demand$1.09
Quadro RTX 600024 GB · on demand$0.69At GPUBank
vCPU, RAM, NVMe and traffic included. All prices
03What the price includes
What appears on the bill besides the GPU.
The same H100 SXM session of 2 h 14 m 03 s, line by line, as each provider describes its billing.
Lambda
2 h 14 m 03 s on H100 SXM
- GPU · $3.99 / h
- $8.91
- vCPU and RAM
- Included
- Storage
- Shown in the console only
- Egress
- Included
- Public IPv4
- Public IP per instance, no charge listed
- Billing granularity
- Per minute
- Minimum charge
- Not published
GPUBank
2 h 14 m 03 s on H100 SXM
- GPU · $1.85 / h
- $4.13
- vCPU and RAM
- Included
- Storage
- Local NVMe included; persistent disk $0.03 per GB-month
- Egress
- Included
- Public IPv4
- Included
- Billing granularity
- Per second, from the health check
- Minimum charge
- None
While stopped
What keeps running on the bill.
Lambda
No stop state, billed until terminated.
GPUBank
Only the persistent disk, $0.03 per GB-month. Compute, IPv4 and scratch cost nothing.
04Where the GPUs run
5 countries against 4, and where they are.
Countries where Lambda says its GPUs run, from its own region list or documentation, against GPUBank's four regions. A pin marks a country, not the exact site.
United States
Japan
India
Germany
Israel
Paris
Frankfurt
Ashburn
Singapore
Lambda5
- United States
Japan
India
Germany
Israel
GPUBank4
Paris- Frankfurt
- Ashburn
Singapore
Lambda · 5 countriesGPUBank · 4 regions
05Paying and signing up
How each one takes your money.
Lambda
Payments and sign-up
Payment methods
Cards
No cryptoMinimum top-up None ($10 card pre-authorization)
To start
- 1Card
Billing modelPostpaid, card on file
GPUBank
Payments and sign-up
Payment methods
A prepaid USD balance, topped up in crypto. Every GPU-second is drawn from it.
Minimum top-up $25
To start
- 1Email and password
- 2Top up in crypto
- 3Launch a GPU
Billing modelPrepaid balance
06Trust and support
Certificates, uptime and who answers.
| Lambda | GPUBank | |
|---|---|---|
| SOC 2 | Type II | Not published |
| ISO 27001 | Yes | Not published |
| HIPAA | Not published | No |
| GDPR DPA | Not published | Published DPA, applies automatically |
| Availability target | No public uptime SLA | 99.9 % monthly, per region |
| Status page | status.lambda.ai | Public status page |
| Support | Tickets and forum; 24/7 on clusters | Tickets from a $25 balance |
07Their site, our site
What you see before you pay.
Both sites as a visitor sees them, without an account: the Lambda pricing page captured on 29 September 2026, and GPUBank's live board. Cookie banners and chat bubbles are closed; nothing else is edited.
- 1Per GPU-hour prices for every instance size are public, in 8x, 4x, 2x and 1x tabs.
- 2Every price is marked plus applicable sales tax, VAT or GST.
- 3Cluster rows lead to Talk to our team, and 1-year terms show no price.
- 1Live stock per region and every price, without an account.
- 2Per-hour and per-second prices, with vCPU, RAM, NVMe, IPv4 and traffic in them.
- 3A machine-readable price feed at /prices.json.
08The spec sheet
27 points, read off both sites.
GPUBank aheadLambda aheadEven
| Criterion | GPUBank | Lambda |
|---|---|---|
| Price and catalogue | ||
| H100 SXM, per GPU-hour | $1.85, all-in | $3.99 (8-GPU), $4.29 (1 GPU) |
| Cheapest GPU-hour | $0.23, RTX 4090 | $0.69, Quadro RTX 6000 |
| GPU models listed | 10 | 14Ahead |
| What the GPU price includesBoth include compute and a local disk in the GPU price. | vCPU, RAM, NVMe, IPv4, traffic | vCPU, RAM, local SSD |
| Billing | ||
| Billing granularity | Per second | Per minute |
| Minimum charge | None | Not published |
| Stopped instance | Disk only | No stop state, billed until terminated |
| Persistent storageDocs quote $0.20 per GiB-month as an example that may not reflect current pricing. | $0.03 per GB-month | Shown in the console only |
| Egress traffic | Included | Included |
| Public IPv4 | Included | Public IP per instance, no charge listed |
| Billing model | Prepaid balance | Postpaid, card on file |
| Payment and sign-up | ||
| Crypto paymentsLambda billing docs list major credit cards; crypto is not mentioned. | Bitcoin, USDT, Ethereum and 12+ more | Not offered: credit cards only |
| Card payments | No | Major credit cardsAhead |
| Minimum deposit | $25 | None ($10 card pre-authorization)Ahead |
| Platform | ||
| API and CLI | REST API, CLI, Python SDK | REST API; no official CLI found |
| Terraform | Official provider | No provider found |
| Serverless GPUs | No | No |
| Managed Kubernetes | No | Managed, on 1-Click ClustersAhead |
| Ready-made templates | 25 | Lambda Stack image |
| Interruptible or spot | Yes, 40 % off | No |
| GPUs per instance | Up to 8 | Up to 8, clusters of 16+Ahead |
| Multi-node interconnect | Not offered | InfiniBand, up to 3,200 Gb/sAhead |
| Network and regions | ||
| Regions | 4 | 16Ahead |
| Countries | France, Germany, US, Singapore | US, Japan, India, Germany, IsraelAhead |
| EU regions | Paris, Frankfurt | One, in Germany |
| Trust and support | ||
| Availability target | 99.9 % monthly | No public uptime SLA |
| Compliance reports | None published | SOC 2 Type II, ISO 27001Ahead |
| Public status page | Yes | status.lambda.ai |
| Support | Tickets from a $25 balance | Tickets and forum; 24/7 on clustersAhead |
09The right fit
Different needs. A clear choice.
What Lambda does better
- Reserved InfiniBand clusters of 16 to 2,000+ GPUs
- B200 SXM6 on demand, GH200 at $2.29 per GPU-hour
- SOC 2 Type II and ISO 27001, 27017, 27701, 22301 stated
- 16 regions in the US, Japan, India, Germany and Israel
- Managed Kubernetes or Slurm on clusters
Choose GPUBank if
- H100 SXM at $1.85 per GPU-hour, against $3.99 to $4.29
- Per-second billing that starts after the health check
- Stop an instance and pay only for its disk
- Pay in Bitcoin, USDT and 12+ other coins, no card
- RTX 4090, RTX 5090, L40S and MI300X, which Lambda does not list
Moving from Lambda takes an afternoon.
Lambda has no egress fees, so moving data out costs nothing on its side. Its docs already advise containerising environments, which makes them portable.
- 1Launch the same GPU on GPUBank with a persistent disk in the region you need.
- 2Copy your filesystem with rsync, or through Lambda's S3 adapter with rclone.
- 3Run the same Docker image or pick the matching template, then stop Lambda instances.
# 1. a GPU and a disk on GPUBank gpubank launch --gpu h100-sxm --region eu-paris --disk data # 2. copy your data from the old machine rsync -azP old-host:/workspace/ root@<new-ip>:/workspace/ # 3. run the same image or template ssh root@<new-ip> 'cd /workspace && ./start.sh'
A few things you might be wondering.
Yes. Lambda lists the H100 SXM at $3.99 per GPU-hour on an 8-GPU instance and $4.29 on a single GPU. GPUBank charges $1.85 per GPU-hour for an H100 SXM, with vCPU, RAM, NVMe, IPv4 and traffic included.
No. Lambda instances can be launched, restarted or terminated, not stopped, and they are billed until terminated. On GPUBank a stopped instance costs only its persistent disk, at $0.03 per GB-month.
Lambda does not mention crypto. Its billing docs list major credit cards and decline prepaid and debit cards. GPUBank is paid from a USD balance topped up in Bitcoin, USDT, Ethereum and more than a dozen other coins.
Lambda. It reserves 1-Click Clusters of 16 to 2,000+ B200 or H100 GPUs on InfiniBand with managed Kubernetes or Slurm. GPUBank rents up to 8 GPUs per instance and does not sell InfiniBand clusters.
Lambda lists one European region, europe-central-1 in Germany, among its 16 regions. GPUBank has two European regions, Paris and Frankfurt, plus Ashburn and Singapore.
Without. Every price on this page excludes VAT and sales tax, for Lambda and for GPUBank, so the two columns compare like for like. Tax is added at checkout according to where the buyer is based.
Sources
Every figure about Lambda was read on its own pages or official documentation. Prices last checked on 29 September 2026.
- 01lambda.ai/pricingAll on-demand and 1-Click Cluster GPU prices (H100 cluster tab read from page HTML)
- 02lambda.ai/instancesPer-minute billing, no egress fees, 1x to 8x instances
- 03lambda.ai/1-click-clustersCluster scale, ISO/SOC 2 claims, managed orchestration
- 04lambda.ai/superclustersGB300 NVL72, HGX B300, Vera Rubin NVL72, 4, 000 to 165, 000+ GPUs
- 05docs.lambda.ai/public-cloud/billing/Per-minute billing, weekly invoices, filesystem billing, credits and refunds
- 06docs.lambda.ai/public-cloud/manage-billing/Credit card only, 10 USD pre-authorization, supported purchase countries
- 07docs.lambda.ai/public-cloud/on-demand/Instance types, images, JupyterLab, regions, filesystem limits
- 08docs.lambda.ai/public-cloud/on-demand/creating-managing-instances/No stop/suspend, new-account quota
- 09docs.lambda.ai/public-cloud/filesystems/Filesystem capacity, no cross-region transfer
- 10docs.lambda.ai/public-cloud/s3-adapter-filesystems/S3 endpoints, regions us-east-3 and us-midwest-2
- 11docs.lambda.ai/public-cloud/importing-exporting-data/Migration paths
- 12docs.lambda.ai/public-cloud/1-click-clusters/InfiniBand fabric, head nodes, local NVMe, reservation and 10-day invoice
- 13docs.lambda.ai/public-cloud/1-click-clusters/support/Support team and response times
- 14docs.lambda.ai/public-cloud/1-click-clusters/security-posture/Single-tenant compute nodes, encryption at rest
- 15docs.lambda.ai/public-cloud/access-security/API keys, roles, audit logs, trust portal
- 16docs.lambda.ai/managed-kubernetes/Managed Kubernetes on 1CC
- 17docs.lambda.ai/api/cloudAPI availability
- 18status.lambda.ai/Status page
What we could not verify (7)
- Current filesystem price per GiB-month (docs example of 0.20 USD may not reflect current pricing; real price shown only in the console)
- Price of 1-year-plus 1-Click Cluster reservations and of Superclusters (GB300 NVL72, HGX B300, Vera Rubin NVL72): not published
- Whether 1CC invoices can be paid by wire/ACH; PayPal and crypto not mentioned anywhere
- HIPAA and GDPR DPA status: trust portal (trust.lambda.ai) returned a bot challenge
- Official CLI, SDK and Terraform provider: none found in docs; registry.terraform.io/providers/lambdal/lambda returned 404
- Free credits or trial: not published
- Per-region GPU availability not published (docs: not every instance type is available in every region)
In short
GPUBank vs Lambda, in facts.
The essentials of this comparison in one table, for quick readers and AI assistants. Prices last checked on 29 September 2026.
Read as Markdown- H100 SXM, per GPU-hour
- Lambda: $3.99 (8-GPU), $4.29 (1 GPU)
GPUBank: $1.85, all-in - Cheapest GPU-hour
- Lambda: $0.69, Quadro RTX 6000
GPUBank: $0.23, RTX 4090 - Billing granularity
- Lambda: Per minute
GPUBank: Per second - Stopped instance
- Lambda: No stop state, billed until terminated
GPUBank: Disk only - Persistent storage
- Lambda: Shown in the console only
GPUBank: $0.03 per GB-month - Egress traffic
- Lambda: Included
GPUBank: Included - Crypto payments
- Lambda: Not offered: credit cards only
GPUBank: Bitcoin, USDT, Ethereum and 12+ more - Card payments
- Lambda: Major credit cards
GPUBank: No - Minimum deposit
- Lambda: None ($10 card pre-authorization)
GPUBank: $25 - GPU models listed
- Lambda: 14
GPUBank: 10 - Regions
- Lambda: 16
GPUBank: 4 - Serverless GPUs
- Lambda: No
GPUBank: No - Availability target
- Lambda: No public uptime SLA
GPUBank: 99.9 % monthly - Compliance reports
- Lambda: SOC 2 Type II, ISO 27001
GPUBank: None published - Support
- Lambda: Tickets and forum; 24/7 on clusters
GPUBank: Tickets from a $25 balance - Headquarters
- Lambda: San Jose, United States
GPUBank: Tallinn, Estonia
How this page was made
Every price and policy of Lambda was read on its own website, documentation or public API on 29 September 2026, with the address of each page listed above. A value we could not read on a primary source is marked "Not published" rather than estimated. GPUBank figures come from this site and its terms. Prices change: check both providers before you buy. Lambda, its logo and its website belong to their owner and are shown only to identify the service compared. GPUBank is not affiliated with Lambda.
