The eval item object

⚠️

Preview — this feature is being built; the design below is the spec.

The eval_item object as the API returns it. Every field below appears in every response that returns this object; fields with no value are null rather than missing.

Example

{
  "id": "evit_01JD24C6E8G0J2L4N6Q8",
  "type": "eval_item",
  "dataset_id": "evds_01JD13B5D7F9H1K3M5P7",
  "input": {
    "ticket": "I was charged twice for order 88123.",
    "customer_tier": "gold"
  },
  "ideal": "Apologise, confirm the duplicate charge, refund within 3 working days, offer a 10% voucher to gold customers.",
  "tags": [
    "billing"
  ],
  "created_at": "2026-11-01T09:00:00Z"
}

Fields

Field	Type	Description
`id`	string	Unique id of this object.
`type`	string	The object type, always `eval_item`.
`dataset_id`	string	The dataset.
`input`	object	Values substituted into the prompt template’s `{{variables}}`.
`ideal`	string or object or null	The reference answer graders compare against. Optional for rubric graders.
`tags`	array	Up to 10 tags; results can be broken down by tag.
`created_at`	string	When it was created, RFC 3339 UTC (some older objects use Unix seconds).

Returned by

eval_grader eval_run

Was this page helpful?