Manage run Grader.

Contract-preserving POST Endpoint for run Grader in the OpenAI API. Model and resource-specific values are supplied by the caller at runtime.

All 288 OpenAI Platform REST endpoints from a fixed MIT-licensed official revision, including typed SSE streaming and Realtime call management. Hub provides documentation and an independent SDK/CLI, but never proxies model, account, or usage data; callers connect directly with local API keys.

Request

operationId: runGrader
Server: https://api.openai.com/v1
Authentication: ApiKeyAuth, AdminApiKeyAuth

Request parameters

No URL, query, or header parameters.

Request body

RunGraderRequestrequired

Content-Type: application/json

Properties: graderitemmodel_sample

Responses

`200`

RunGraderResponseapplication/json

Successful supplier response.

Properties: metadatamodel_grader_token_usage_per_modelrewardsub_rewards

{
  "metadata": {
    "errors": {
      "formula_parse_error": false,
      "invalid_variable_error": false,
      "model_grader_parse_error": false,
      "model_grader_refusal_error": false,
      "model_grader_server_error": false,
      "other_error": false,
      "python_grader_runtime_error": false
    },
    "execution_time": 0,
    "scores": {}
  },
  "model_grader_token_usage_per_model": {},
  "reward": 0,
  "sub_rewards": {}
}