Lambda

What it does

Lambda offers cloud supercomputers for AI training and inference, using NVIDIA GPUs such as the B200, H100, A100, and GH200. Users can launch on-demand instances in minutes or reserve 1-Click Clusters and Superclusters for larger jobs. It targets AI developers, enterprises, and research teams that need scalable GPU compute.

Lambda provides cloud supercomputers for AI training and inference. The homepage says you can launch a GPU instance and describes single-tenant infrastructure with SOC 2 Type II certification. The pricing page lists on-demand instances with NVIDIA B200, H100, A100, and GH200 GPUs, priced per GPU per hour. It also lists 1-Click Clusters from 16 to 2,000+ NVIDIA B200 or H100 GPUs, with reserved capacity quoted on request. The site does not state whether a free plan, API, or open source license is available. Before committing, check current GPU availability, the exact per-hour rate for your chosen instance, and whether your region and data rules fit the single-tenant model.