Skip to main content
GET
/
vault
/
{id}
/
objects
/
{objectId}
/
ocr-words
Get OCR word bounding boxes
curl --request GET \
  --url https://api.case.dev/vault/{id}/objects/{objectId}/ocr-words \
  --header 'Authorization: Bearer <token>'
{
  "objectId": "obj_abc123",
  "pageCount": 5,
  "totalWords": 2500,
  "pages": [
    {
      "page": 1,
      "words": [
        {
          "text": "The",
          "bbox": [
            0.12,
            0.71,
            0.15,
            0.75
          ],
          "confidence": 0.98,
          "wordIndex": 0
        },
        {
          "text": "witness",
          "bbox": [
            0.16,
            0.71,
            0.28,
            0.75
          ],
          "confidence": 0.99,
          "wordIndex": 1
        }
      ]
    }
  ],
  "createdAt": "2024-01-15T10:30:00Z"
}

Authorizations

​
Authorization
string
header
required

API key starting with sk_case_

Path Parameters

​
id
string
required

The vault ID

​
objectId
string
required

The object ID

Query Parameters

​
page
integer

Filter to a specific page number (1-indexed). If omitted, returns all pages.

​
wordStart
integer

Filter to words starting at this index (inclusive). Useful for retrieving words for a specific chunk.

​
wordEnd
integer

Filter to words ending at this index (inclusive). Useful for retrieving words for a specific chunk.

Response

Successfully retrieved OCR word data

​
objectId
string

The object ID

​
pageCount
integer

Total number of pages in the document

​
totalWords
integer

Total number of words extracted from the document

​
pages
object[]

Per-page word data with bounding boxes

​
createdAt
string<date-time>

When the OCR data was extracted