The eval dataset object

⚠️

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

The eval_dataset 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": "evds_01JD13B5D7F9H1K3M5P7",
  "type": "eval_dataset",
  "name": "support-replies-v3",
  "description": "200 real tickets with approved answers",
  "item_count": 200,
  "input_schema": {
    "type": "object",
    "properties": {
      "ticket": {
        "type": "string"
      },
      "customer_tier": {
        "type": "string"
      }
    },
    "required": [
      "ticket"
    ]
  },
  "created_at": "2026-11-01T09:00:00Z",
  "updated_at": "2026-11-03T12:00:00Z"
}

Fields

Field	Type	Description
`id`	string	Starts with `evds_`.
`type`	string	The object type, always `eval_dataset`.
`name`	string	Name shown in the Console and in reports.
`description`	string	Free-text description.
`item_count`	integer	Items in the dataset, up to 50,000.
`input_schema`	object or null	Optional JSON Schema every item’s `input` must match. Catches malformed items at upload time.
`created_at`	string	When it was created, RFC 3339 UTC (some older objects use Unix seconds).
`updated_at`	string	When it last changed.

Returned by

environment eval_grader

Was this page helpful?