RunPod is a managed GPU cloud and serverless inference platform offering on-demand and persistent GPU Pods, autoscaling Serverless endpoints, network volumes, container templates, and a REST + GraphQL control plane for provisioning H100, H200, B200, A100, L40S, and consumer RTX GPUs. - View it on GitHub