ModelRefInfo¶

Model reference with rate limit information from the passthrough endpoint

Type: object

Properties¶

Property	Type	Required	Description
`name`	string	Yes	Name of the MaaSModelRef Example: `granite-3-8b-instruct`
`display_name`	string	No	Human-readable display name of the model (from openshift.io/display-name annotation) Example: `Granite 3 8B Instruct`
`source`	string	No	Model source identifier (internal or external) Example: `internal`
`description`	string	No	Human-readable description of the model Example: `Granite 3 8B Instruct is a large language model that is used for advanced tasks.`
`namespace`	string	No	Namespace where the MaaSModelRef lives Example: `maas-models`
`token_rate_limits`	array of object	No
`token_rate_limits[].limit`	integer (int64)	Yes	Maximum number of tokens allowed Example: `100000`
`token_rate_limits[].window`	string	Yes	Time window (e.g., 1m, 1h, 24h) Example: `24h`
`billing_rate`	object	No
`billing_rate.per_token`	string	No	Cost per token Example: `0.001`

Example¶

{
  "name": "granite-3-8b-instruct",
  "display_name": "Granite 3 8B Instruct",
  "source": "internal",
  "description": "Granite 3 8B Instruct is a large language model that is used for advanced tasks.",
  "namespace": "maas-models",
  "token_rate_limits": [
    {
      "limit": 100000,
      "window": "24h"
    }
  ],
  "billing_rate": {
    "per_token": "0.001"
  }
}