PDF to JSON

Convert a PDF to structured JSON: text, tables, form fields and metadata

A PDF is made for reading, not for programs. PDF to JSON turns a PDF into one JSON file that code can work with: metadata, the text of every page, the headings, paragraphs, lists and tables found in the layout, and the values of any form fields. It is the step before loading a document into a database, a search index or an AI prompt.

Upload the PDF and pick what you need. Structure gives you everything: a source block with title, author, page count and dates; a pages array with the text and size of each page; a blocks array in reading order of headings, paragraphs, lists and tables; and a tables array with every table as typed rows, tagged with its page. Tables only returns just the tables, the fast route for invoices and statements. Page text only returns the raw text per page. Numbers inside tables come back as numbers, so a total is a value rather than a string.

The output follows one documented shape, described in the JSON guide, so a script written today keeps working. In Structure mode a scanned PDF is recognised first and the extraction block says so; a password-protected file has to be unlocked first. For the reverse direction use JSON to PDF, for a spreadsheet use PDF to Excel, and for prose use PDF to Markdown.