Retrieve a model · OpenGptApi

Overview

One logical model may have multiple route rows. Read each row's closed capability set. Only exact published and routable combinations may be submitted; omitted 1K, 2K, 4K, or ratio combinations must not be inferred.

Authentication: No authentication

Request

Path parameters

Name	Type	Description
`model_id`required	string	Published logical model ID.

Response

JSON endpoints use the standard response envelope. HTTP status represents the request outcome; business errors are not disguised as HTTP 200.

HTTP	Meaning
`200`	Request succeeded
`404`	Resource not found or signature invalid
`422`	Request-structure validation failed
`500`	Internal service error

Complete success example (optional fields vary by state)

{
  "code": 0,
  "message": "success",
  "data": {
    "model_id": "gpt-image-2",
    "company": "OpenAI",
    "modality": "image",
    "lifecycle_status": "active",
    "availability_status": "available",
    "capabilities": [],
    "verified_capabilities": []
  },
  "request_id": "8f25b42f-2ed1-4ad7-99de-61d7f779bc7b",
  "timestamp": "2026-08-10T00:00:00Z"
}

Important boundaries

Use the public price_matrix for customer pricing, never supplier acquisition cost.
All three Nano Banana models explicitly return custom_aspect_ratio_supported=false and custom_pixel_size_supported=false; submit only discrete tier/ratio pairs listed by a route.
Arbitrary WIDTHxHEIGHT may be attempted only when a current routable route explicitly returns parameters.size.custom_size_constraints. Without that field, custom pixel sizes are unsupported.

Open the machine-readable OpenAPI