Question 1

How is a PDF to markdown conversion priced?

Accepted Answer

Per page, at one of two rates. Pages with a real text layer cost one credit per ten pages. Pages that need OCR cost one credit each. Which is which is decided by classifying the document before anything is charged, so a mixed file is billed at both rates rather than at the worse one.

Question 2

Do I get chunks for a vector database?

Accepted Answer

Yes. Pass chunk=true and the response includes heading-aware chunks alongside the markdown, with a size target you set. Chunking costs no extra credits.

Question 3

Can I force or disable OCR?

Accepted Answer

Yes. The ocr parameter takes auto, force or never. Auto is the classifier, force runs OCR on every page, and never fails rather than silently returning an empty document for a scan.

Question 4

Is my document stored?

Accepted Answer

No. It is written to a memory-backed workspace, parsed, and the workspace is removed before the response completes. There is no object storage in the request path and no model is trained on anything you send.

PDF to Markdown