Models

Prefix: /api/v1/model-catalog

See Overview for authentication and pagination.

id on a catalog row is the catalog listing id. Use that id for the detail call and for Playground Try (configModelId).

Endpoint summary

Method	Path
GET	`/model-catalog`
GET	`/model-catalog/{listingId}`

List models

GET /api/v1/model-catalog?page=0&size=20&sortBy=popularity
Authorization: Bearer YOUR_API_TOKEN

Query	Default	Description
`page`	`0`	0-based page index
`size`	`20`	Page size, clamped to 1–100
`search`		Case-insensitive match across the model search text
`sortBy`	`popularity`	`popularity`, `name`, `author`, `modality`, or `price`. Any other value sorts as `popularity`
`direction`	`DESC` when `sortBy` is `popularity`	`ASC` or `DESC`
`authors`		Comma-separated author UUIDs
`modalities`		Comma-separated modality UUIDs
`compliance`		Comma-separated compliance UUIDs
`categories`		Comma-separated category UUIDs
`capabilities`		Comma-separated capability UUIDs
`quantizations`		Comma-separated quantization UUIDs
`gpus`		Comma-separated GPU SKU values
`costMin`		Minimum average price per million tokens
`costMax`		Maximum average price per million tokens
`latency`		`fast` (under 500 ms), `medium` (500 ms up to under 2000 ms), or `slow` (2000 ms or more). Rows with no latency are omitted when this is set

Response 200 — StandardPageResponse of catalog rows. Fields that have no value are returned as null. The row also includes ids and labels for license, region, framework, serving mode, compliance, and modalities.

{
  "items": [
    {
      "id": "aaaaaaaa-bbbb-cccc-dddd-eeeeeeeeeeee",
      "name": "gpt-4o",
      "authorId": "11111111-1111-1111-1111-111111111111",
      "author": "OpenAI",
      "description": "General purpose model",
      "configModelId": "22222222-2222-2222-2222-222222222222",
      "configModelName": "gpt-4o",
      "catalogModelId": "33333333-3333-3333-3333-333333333333",
      "modalityId": "44444444-4444-4444-4444-444444444444",
      "modalityLabel": "Text",
      "inputCostPerToken": 0.0000025,
      "outputCostPerToken": 0.00001,
      "cachedInputCostPerToken": 0.00000125,
      "pricingKind": "SIMPLE",
      "contextWindow": 128000,
      "latencyMs": 420,
      "ttftMs": 180,
      "throughputTokensSec": 90,
      "qualityScore": 0.91,
      "listingStatus": "Active",
      "capabilityLabels": ["Reasoning", "Code"],
      "inputModalityLabels": ["Text"],
      "outputModalityLabels": ["Text"]
    }
  ],
  "page": 0,
  "size": 20,
  "totalElements": 1,
  "totalPages": 1
}

pricingKind is SIMPLE, ADVANCED, or NONE.

Get one model

GET /api/v1/model-catalog/{listingId}
Authorization: Bearer YOUR_API_TOKEN

listingId is the catalog row id from the list.

Response 200

{
  "listing": {
    "id": "aaaaaaaa-bbbb-cccc-dddd-eeeeeeeeeeee",
    "name": "gpt-4o",
    "author": "OpenAI",
    "inputCostPerToken": 0.0000025,
    "outputCostPerToken": 0.00001,
    "pricingKind": "SIMPLE"
  },
  "pricingKind": "SIMPLE",
  "serviceId": "55555555-5555-5555-5555-555555555555",
  "matrixRates": [
    {
      "matrixRowId": "66666666-6666-6666-6666-666666666666",
      "dimensionName": "Input_Tokens",
      "unitRatePerToken": 0.0000025,
      "unitRatePerMillion": 2.5,
      "priceMethod": "QUANTITY",
      "pricingKind": "SIMPLE",
      "unitRate": 0.0000025
    }
  ],
  "vendor": {},
  "model": {},
  "quantization": {},
  "configModel": {}
}

listing is the same catalog row as the list. matrixRates is the service rate card. vendor, model, quantization, and configModel are the raw records linked to that listing.

A missing or governance-hidden listing returns 400:

{
  "status": 400,
  "error": "Validation Error",
  "message": "Listing not found: aaaaaaaa-bbbb-cccc-dddd-eeeeeeeeeeee"
}

Models

Models

Endpoint summary

List models

Get one model

On this page