gpu.aiDocs
CH·02CLI reference

gpu community machines register

Register a machine and mint its one-time enrollment token

Register a machine and mint its one-time enrollment token

Synopsis

Registers a self-declared machine under your Community Cloud supplier role and mints a one-time enrollment token. All flags are optional — a bare register creates a pending_verification placeholder with no declared specs, and only the flags you explicitly pass are sent to the server.

The enrollment token is displayed exactly once and cannot be retrieved
again. Save it immediately — it is consumed by the machine-agent installer
(Phase 56) to enroll the host. If you lose it, register a new machine to mint
a fresh token.
gpu community machines register [flags]

Examples

# Placeholder registration (fill in specs later)
gpu community machines register

# Declare specs up front (defaults to on-demand capacity)
gpu community machines register \
  --gpu-type H100 --gpu-count 8 --vram-gb 80 \
  --price-per-hour 1.85 --region us-west

# List the machine as cheaper, interruptible spot capacity instead
gpu community machines register \
  --gpu-type H100 --gpu-count 8 --price-per-hour 1.20 --tier spot

Options

      --gpu-type string          GPU model, e.g. H100 (optional)
      --gpu-count int            Number of GPUs, 1-64 (optional)
      --vram-gb int              VRAM per GPU in GB, 1-2048 (optional)
      --price-per-hour float     Asking price per hour in USD (optional)
      --region string            Machine region label (optional)
      --tier string              Rental model: on_demand (default) or spot (optional)
  -h, --help                     help for register

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

← The gpu CLI