Lambda: Pricing, Features & Alternatives
Lambda is the training-focused GPU cloud. It offers on-demand instances with ML frameworks preinstalled, and 1-Click Clusters of 16 to 256+ interconnected GPUs reserved for two weeks up to a year. On-demand pricing per GPU-hour as of September 2026: an H100 SXM is $3.99–4.29, an H100 PCIe $3.29, a GH200 $2.29, an A100 from $1.99, and a B200 $6.69–6.99. Clusters cost more per GPU-hour ($5.54–6.16 for H100, $8.87–9.86 for B200) because they come with the high-bandwidth interconnect multi-node training needs. Longer commitments are quoted.
Category
GPU Cloud & Inference
Pricing
Pay-as-you-go
Free tier
No
Best for
GPU Cloud & Inference — gpu, cloud
Lambda Pricing Plans (2026)
| Plan | Price |
|---|---|
| On-demand instancesPopular | H100 PCIe $3.29 / GPU-hour — A100 from $1.99 |
| 1-Click Clusters | H100 from $5.54 / GPU-hour (2-week to 1-year terms) |
Pricing summary: Usage-based. Always confirm current pricing on the official site.
Key Lambda Features
- H100 SXM $3.99–4.29 / GPU-hour
- GH200 $2.29 / GPU-hour
- B200 $6.69–6.99 / GPU-hour
Pros
- +Built for training, with multi-node clusters on short terms
- +ML stack preinstalled, so there is little setup
- +Among the cheapest GH200 and A100 hours
Cons
- −On-demand capacity is often sold out
- −No scale-to-zero serverless inference
- −H100 on-demand costs more than Runpod
Best Lambda Alternatives
Compare allLambda Compared
Lambda FAQ
What is Lambda used for?⌄
Lambda is a gpu cloud & inference tool. A GPU cloud built for AI training: on-demand H100 and B200 instances and 1-Click Clusters for multi-node jobs.
Is Lambda free?⌄
Lambda does not offer a permanent free tier.
How much does Lambda cost?⌄
Lambda uses usage-based (pay-as-you-go) pricing — see the official pricing page for current rates.
What are the best Lambda alternatives?⌄
Popular Lambda alternatives include Runpod, Replicate. Compare pricing and features on our Lambda alternatives page.
Not sure if Lambda fits your stack?
Get a free, AI-powered tech stack tailored to your budget, app type, and team size — including the best gpu cloud & inference pick for you.
Build my stack free