Skip to content

rp pod

On-demand GPU and CPU pod lifecycle. A pod is a single container billed per second, built from an image (or from a template's container config) and addressed by id. Every pod is either a GPU pod or a CPU pod, never both. Storage is one host-local persistent volume or one network volume, and the choice is fixed at create time.

rp pod <verb> [flags]

Commands

  • rp pod create — Create a pod from an image, optionally seeded by a template.
  • rp pod list — List your pods as a table: id, name, image, status, cost.
  • rp pod get — Show one pod's full record, including runtime ports and utilisation.
  • rp pod update — Change a pod's configuration in place, restarting it.
  • rp pod delete — Terminate a pod permanently.
  • rp pod start — Start a stopped pod.
  • rp pod stop — Stop a running pod, keeping its disks.
  • rp pod reset — Deprecated: alias for restart — v2 removed the reset action.
  • rp pod restart — Restart a running pod.
  • rp pod logs — Stream a pod's container and system logs live.