Unsiloed AI

Creates a parse job for a PDF, DOCX, JPEG, PNG, WEBP or PPTX document sent by URL, and returns the job id at once. Other formats are converted to PDF first. Poll GET /parse/{job_id} for the ordered chunks with Markdown, HTML, OCR text, bounding boxes and confidence per segment. Options select the layout model, OCR engine, page range, output fields and exports. Excel workbooks go to POST /parse/excel. Usage is the page count of the document.

Creates an extraction job for a PDF given by URL, with a JSON Schema of the fields to pull out, and returns the job id at once. Poll GET /extract/{job_id} for the typed values, each with a confidence score and, when citations are on, the page and bounding box it came from. Model tiers trade speed for thoroughness; high_accuracy_mode adds a verification pass. Usage is the page count of the PDF.

Creates a classification job for a PDF or image (JPG, PNG, GIF, WebP, BMP, TIFF) given by URL, with the list of candidate category names, and returns the job id at once. Poll GET /classify/{job_id} for the dominant category, its confidence, every category present with the pages that voted for it, and per-page results. Documents longer than four pages are classified from their first four. Usage is the page count of the document.