Skip to main content
GET
Get the status or result of a parse or extraction job
Retrieve a job and read its status, error, and result. Request server-sent events to follow progress while it runs. Start with Parse your first document.

Authorizations

unstructured-api-key
string
header
required

API keys must be sent in this header.

Path Parameters

jobId
string
required

Identifies one piece of work, whichever path submitted it. The same id the blocking call returns.

Example:

"9eb6c914-02a5-4c5d-8490-a06476946a38"

Query Parameters

output
enum<string>
default:markdown

Shape of the parsed output.

Available options:
markdown,
elements
include
string[]

Optional list of projection fields to include in elements output. Allowed values are 'coordinates' and 'table_html'. Only valid when output='elements'.

Response

The job's state, carrying parsed content as soon as it is available. As an event stream, progress events then one terminal result or error event.

A job's state, carrying parsed content as soon as it is available.

id
string
required

Identifies one piece of work, whichever path submitted it. The same id the blocking call returns.

Example:

"9eb6c914-02a5-4c5d-8490-a06476946a38"

status
enum<string>
required

Terminal values match the blocking call's own status values.

Available options:
queued,
processing,
failed,
completed,
completed_with_warnings,
cancelled
file_id
string | null
required

The authoritative source/input file ID. This is the ID accepted by GET /api/v2/upload/{fileId}, never a job ID, result artifact ID, storage key, or signed URL. Null for legacy or unassociated jobs.

request_source
enum<string> | null
required

Where the job was created, when known.

Available options:
transform-ui,
transform-api
filename
string | null

The document's original uploaded filename.

cancellation_requested
boolean
default:false

True after cancellation has been requested for this job.

source
object | null

The associated original uploaded source file, when known.

error
object | null

A machine-readable code and an actionable message.

configuration
object | null

The settings this job ran with, available in every status. Null for extraction jobs, which run no partitioner of their own.

result
object | null

Null until parsed content is available.