Skip to main content
GET

Authorizations

Authorization
string
header
required

API key starting with sk_case_

Path Parameters

id
string
required

The OCR job ID returned from the create OCR endpoint

Query Parameters

include_text
enum<string>

Include full OCR text in completed responses (default: false)

Available options:
true,
false

Response

OCR job status and results

id
string
required

OCR job ID

status
enum<string>
required

Current job status

Available options:
pending,
processing,
completed,
failed
created_at
string<date-time>
required

Job creation timestamp

error
string | null

Safe human-readable failure message; null when no failure is recorded.

error_code
string | null

Stable failure identifier, such as source_access_denied, invalid_pdf, inference_failed, chunks_failed, or result_upload_failed. Historic failures may be unclassified.

error_stage
enum<string> | null

Operation that failed.

Available options:
dispatch,
download,
document_validation,
chunking,
engine_init,
inference,
feature_processing,
result_write,
chunk_processing,
result_assembly,
result_upload
retryable
boolean | null

Whether retrying the same document may help. Unknown failures retain bounded retry behavior.

failed_chunks
object[] | null

Bounded failure samples. Page ranges are one-based and inclusive; chunks_failed and chunk_count retain exact totals.

Maximum array length: 20
chunks_failed
integer

Exact number of failed OCR chunks.

Required range: x >= 0
chunk_count
integer

Total number of OCR chunks.

Required range: x >= 0
page_count
integer

Number of pages processed

text
string

Extracted text content when completed and include_text=true

metadata
object

Additional processing metadata

completed_at
string<date-time>

Job completion timestamp