Create Splitting Job

POST 
/splittings/jobs

Start a document splitting job.

Creates a background job that detects sub-document boundaries within a multi-page document. Only page-based documents (PDF and Word) can be split. Returns a job ID that can be used to poll for the detected boundaries via GET /splittings/jobs/{job_id}.

Rules: pass rules to constrain detection to known splitting types. Omit rules (or pass an empty list) for class-agnostic detection: boundaries are detected without predefined types and each splitting's name carries a model-detected label instead of a rule name.

Indexing: the document must be indexed before splitting runs. If it is still indexing when the request arrives, the job waits for indexing to finish, but may fail if indexing does not complete within the processing window. Upload documents in advance to minimize latency.

Request

Responses

201

Splitting job created successfully

401

Unauthorized

403

Forbidden

404

Not Found

422

Unprocessable Entity