Models

This is one of the safest read-only endpoints to call first during integration.

GET /v1/models

Overview

Item	Description
Route	`GET /v1/models`
Auth	`Authorization: Bearer sk_aibubu_xxx`
Request body	None
Response	`200 OK` + model list

Typical use cases

Fetch the available model list at startup
Verify that a target model is enabled in the current environment
Build your own model selector in a dashboard or config UI

Authorizations

Authorization: Bearer sk_aibubu_your_api_key

Response

A successful response usually returns a model list object.

{
  "object": "list",
  "data": [
    {
      "id": "gpt-image-2",
      "name": "gpt-image-2",
      "region": "all",
      "family": "image",
      "description": "General image generation",
      "capabilities": ["image_generation"],
      "pricing": {"unit": "image", "display": "per image", "credits_text": "use /v1/quote"},
      "schema": {"fields": [{"name": "resolution", "type": "string", "required": false, "default": "1K", "options": ["1K", "2K", "4K"]}]},
      "ui_hints": [],
      "enabled": true
    }
  ]
}

Response fields

Field	Type	Description
`object`	`string`	List object type, usually `list`
`data`	`array`	Currently available models
`data[].id`	`string`	Model ID to pass in later requests
`data[].family`	`string`	Model family: `text`, `image`, or `video`
`data[].schema.fields`	`array`	Dynamic fields with names, types, required/default values, enum `options`, and limits
`data[].pricing`	`object`	Display metadata; use `/v1/quote` for accurate credits
`data[].enabled`	`boolean`	Whether the model is currently enabled

Integration notes

Do not infer capabilities from model names; read family and schema.fields
There is no public /v1/models/{model}/schema; schema data is embedded in each list item
For parameter differences across models, see Supported Models and Parameters
For text models, also use schema.fields; do not assume fields such as reasoning or temperature are supported
If your frontend shows a dynamic model list, prefer refreshing it instead of hardcoding it

Current China-region public models

Public text models currently include GPT/GLM, Claude, Gemini, Grok, and DeepSeek families. Video models include the Seedance, Kling, and Gemini Omni families. Availability changes with live routes, region, and API-key permissions; use this endpoint as the source of truth.