generated: '2026-07-21'
method: searched
source: https://docs.w.ai/w.ai-cli-guide
docs: https://docs.w.ai/w.ai-cli-guide
name: wai
description: >-
  First-party w.ai command-line client. Used to log in, contribute a machine's
  idle GPU compute to the network, and manage Auth API keys. Also distributed as
  optimized Docker images (wdotai/wai) for headless/containerized workers.
install:
- platform: linux
  method: shell
  command: 'curl -fsSL https://app.w.ai/install.sh | bash'
- platform: macos
  method: shell
  command: 'curl -fsSL https://app.w.ai/install.sh | bash'
- platform: windows
  method: powershell
  command: 'iwr -useb https://app.w.ai/install.ps1 | iex'
- platform: docker
  method: docker
  command: docker pull wdotai/wai:latest
  image: https://hub.docker.com/r/wdotai/wai
env:
- name: W_AI_API_KEY
  description: Auth API key for headless environments (created in the dashboard).
commands:
- name: login
  description: Authenticate on a device with a browser/desktop environment.
- name: run
  description: Start contributing GPU compute to the network. `-g 0 1` limits to
    specific NVIDIA GPU IDs.
- name: key list
  description: List generated API keys.
- name: revoke
  description: Revoke an API key by token (`wai revoke <token>`).
- name: key help
  description: Show key-management help.
- name: help
  description: Show an overview of the commands.
flows:
- name: Headless worker
  steps:
  - Create an Auth API key at https://app.w.ai/dashboard
  - export W_AI_API_KEY=your_key_here
  - wai run