Skip to main content
GET
/
v3
/
evals
/
datasets
/
{dataset_id}
/
schema
Get dataset schema
curl --request GET \
  --url https://api.example.com/v3/evals/datasets/{dataset_id}/schema \
  --header 'Authorization: Bearer <token>'
{
  "data": {
    "total_items": 500,
    "sampled": false,
    "inputs": {
      "question": {
        "count": 500,
        "data_type": "STRING"
      },
      "context": {
        "count": 423,
        "data_type": "STRING"
      }
    },
    "expected_outputs": {
      "answer": {
        "count": 500,
        "data_type": "STRING"
      }
    },
    "metadata": {
      "trace_id": {
        "count": 500,
        "data_type": "STRING"
      },
      "model": {
        "count": 500,
        "data_type": "STRING"
      }
    },
    "extras": {
      "cost_usd": {
        "count": 120,
        "data_type": "FLOAT"
      }
    }
  },
  "api_version": "3.0",
  "kind": "NORMAL"
}

Authorizations

​
Authorization
string
header
required

Bearer authentication header of the form Bearer <token>, where <token> is your auth token.

Path Parameters

​
dataset_id
string<uuid>
required

Dataset UUID

Response

Dataset schema per bucket

Response object for standard API responses.

​
api_version
enum<string>
default:3.0

API version.

Available options:
2.0,
3.0
​
kind
enum<string>
default:NORMAL

Type of response, indicating a normal response.

Available options:
NORMAL
​
data
object

Schema of existing dataset items. Each bucket maps field keys to their coverage count and inferred type. Empty objects indicate no items exist for that bucket or no keys were found within it.