Inventory
Query available compute resources, their specifications, pricing, and current availability. This endpoint does not require authentication.
List Inventory
List available compute resources, optionally filtered by type or GPU availability.
Endpoint
GET /tha/v2/inventory
Query Parameters
| Field | Type | Description |
|---|---|---|
type | string | Filter by resource type: rental, serverless, or storage. |
gpu | boolean | Filter by GPU availability. Set to true to only return GPU resources. |
Example Request
curl -X GET "https://api.targon.com/tha/v2/inventory?type=rental&gpu=true"
Response
Returns an array of inventory objects.
Each inventory object:
| Field | Type | Description |
|---|---|---|
name | string | Resource identifier (e.g. h200-small). |
display_name | string | Human-readable name. |
description | string | Description of the resource. |
type | string | Resource type (rental, serverless, storage). |
gpu | boolean | Whether this resource includes a GPU. |
spec | object | Hardware specifications. |
cost_per_hour | number | Hourly cost. |
available | integer | Number of units currently available. |
The spec object:
| Field | Type | Description |
|---|---|---|
gpu_type | string | GPU type (e.g. "H200"). |
gpu_count | integer | Number of GPUs. |
vcpu | integer | Virtual CPU cores. |
memory | integer | Memory in MB. |
storage | integer | Storage in MB. |
Example Response
[
{
"name": "h200-small",
"display_name": "NVIDIA H200 - Small",
"description": "Single H200 GPU instance",
"type": "rental",
"gpu": true,
"spec": {
"gpu_type": "H200",
"gpu_count": 1,
"vcpu": 12,
"memory": 115000,
"storage": 0
},
"cost_per_hour": 2.49,
"available": 5
}
]
Errors
| Status Code | Description |
|---|---|
| 400 - Bad Request | Invalid query parameter (unrecognized type or invalid gpu value). |
| 500 - Internal Server Error | Server error. |