gpu instances ssh

SSH into a running GPU instance

Synopsis

Connects to a running GPU instance via the FRP tunnel. Resolves the host:port from the API, then execs ssh with StrictHostKeyChecking=no etc. so first-connect doesn't prompt. Anything after -- is forwarded to ssh (e.g., -L 8888:localhost:8888 for port-forwarding).

gpu instances ssh <instance-id> [-- ssh-extra-args...] [flags]

Examples

# Basic: connect, waiting up to 10 minutes for status=running
gpu instances ssh gpu-abc123

# Forward a local port to a service on the instance
gpu instances ssh gpu-abc123 -- -L 8888:localhost:8888

# Print the equivalent ssh command without executing it
gpu instances ssh gpu-abc123 --print

# Don't wait — fail fast if not yet running
gpu instances ssh gpu-abc123 --wait=false

Options

  -h, --help                      help for ssh
      --print                     Print the equivalent ssh command instead of executing it
      --wait                      Wait for instance to become running before SSH-ing (default true)
      --wait-timeout duration     How long to wait for status=running (default 10m0s)

Options inherited from parent commands

      --api-base string   API base URL (override with GPUAI_API_BASE env) (default "https://api.gpu.ai/v1")
      --debug             Enable debug logging to stderr
  -o, --output string     Output format: table|json (default table on TTY, json otherwise)

SEE ALSO

• gpu instances - Manage GPU instances