generated: '2026-07-21'
method: searched
source: https://run-ai-docs.nvidia.com/saas/reference/cli/runai.md
docs: https://run-ai-docs.nvidia.com/saas/reference/cli.md
name: runai
description: >-
  The NVIDIA Run:ai Researcher Command Line Interface (runai) — a kubectl-adjacent
  CLI for submitting and managing AI workloads, assets, and cluster resources
  against the Run:ai control plane.
install:
  docs: https://run-ai-docs.nvidia.com/saas/reference/cli/install-cli.md
  self_update: runai upgrade
commands:
- name: auth
  description: Authentication related commands
- name: login
  description: Login to the control plane
- name: logout
  description: Logout of the system
- name: whoami
  description: Show the current logged in user
- name: cluster
  description: Cluster management
- name: node
  description: Node management
- name: nodepool
  description: Node pool management
- name: project
  description: Project management
- name: department
  description: Department management
- name: workload
  description: Workload management
- name: workload-type
  description: Workload type management
- name: workspace
  description: Workspace management
- name: training
  description: Training management
- name: inference
  description: Inference management
- name: template
  description: Workload template management
- name: compute
  description: Compute resource asset management
- name: datasource
  description: Data Source asset management
- name: environment
  description: Environment asset management
- name: pvc
  description: PVC management
- name: my-credential
  description: Manage personal credentials for the current user
- name: kubeconfig
  description: Kubeconfig management
- name: config
  description: Configuration management
- name: diagnostics
  description: System diagnostics and troubleshooting tools
- name: report
  description: '[Experimental] report management'
- name: version
  description: Show the current CLI version
- name: upgrade
  description: Upgrade the CLI to the latest version
framework_aliases:
- pytorch
- tensorflow
- jax
- mpi
- xgboost