Sign in
InstanceTypeService

Get an inference instance type

Retrieves the GPU resources, pricing, regional availability, and best-effort capacity headroom for one inference instance type.

get/public/inference-instance-types/{id}

Path parameters

idstring required

Resource identifier.

Response

OK

idstring required

Stable hardware instance type identifier used by deployment configs.

namestring required

Human-readable instance type name.

descriptionstring required

Human-readable summary of the hardware configuration.

gpuTypestring required

GPU accelerator model, such as H100 or B200.

gpuCountinteger required

Number of GPUs in one replica of this instance type.

gpuMemoryGibinteger required

Memory available on each GPU, in GiB.

priceCentsPerHourinteger required

On-demand price for one running replica, in US cents per hour.

Changes

Changed in 1 of the 14 revisions of this API.1