Resize a GPU VM
Moves a GPU VM to another plan with the same GPU: send the target plan’s cpu, ram_mb, disk_gb and billing_catalog (see the request body). The root disk can grow, never shrink. The same precheck the portal runs is applied first. Asynchronous: the response carries operation_id; poll GET /compute/operations/{operation_id} until status is succeeded, failed, cancelled or timed_out. If your client times out, read the status; never resubmit the request. Requires scope: vm.write.
Authentication
IBEE platform API token. Production tokens start with ibee_prod_key_; development tokens start with ibee_dev_key_. Use a token only with its matching gateway environment.
Path parameters
Headers
Query parameters
The positive numeric workspace ID to scope this request to. Zero, negative, and non-numeric values are rejected.
Request
Copy the selected plan’s billing_catalog, choose one supported HOURLY or MONTHLY entry from its billing_options, and merge that complete entry into the copy. Send the result here, including the canonical price_currency. Hourly requests also require an explicit selected term. Do not invent prices, drop commitment fields from the selected option, or send only billing_interval and committed. Admission validates the selection against the server’s current catalog. Amounts remain server-authoritative.
Response
Always accepted.
When the operation was accepted (ISO 8601).
