Skip to main content
Send a file to POST /parse as raw bytes, without multipart encoding or a JSON wrapper. Files can be up to 50 MiB. For a public document URL, use Scrape documents.

Convert a file

Read markdown for the text and the returned type for the detected format. Request includeLinks, includeImages, shortenBase64Images, or useMainContentOnly as query options when they apply to your input.

File types and OCR

Supply extension when the input type needs a hint; see supported formats. PDFs and OCR covers scanned pages, images, and page ranges. An oversized upload returns 413 and an unsupported format returns 415. A scanned PDF with OCR disabled returns 400 PDF_IMAGES_ONLY; retry it with OCR enabled if text recovery is needed. The Parse reference lists all options and errors.