Choose export detail
Set text-position and metadata options before selecting the file. These settings determine which additional fields the conversion includes.
This export describes what the PDF text extractor can read, rather than interpreting the document as an invoice, database or spreadsheet. Each page has a page number, dimensions, rotation and text. Optional item details help inspect coordinates and font identifiers, but they are not semantic labels or guaranteed reading order.
Set text-position and metadata options before selecting the file. These settings determine which additional fields the conversion includes.
Choose a text-based document. Scanned page images do not gain recognized words in this workflow.
Review page text and item positions, especially on rotated or multi-column pages. The report includes source name and size locally.
Save the JSON and parse its fylvexa.pdf-json.v1 structure in your own workflow. Check required fields and test against representative documents.
Accepts one PDF up to 150 MiB without an explicit page-count cap. Output includes all pages and optional standard metadata information.
No OCR, invoice-field extraction or table-cell inference is performed. Positions describe PDF text items and may need coordinate interpretation. Metadata retrieval can return null when unavailable.
Check whether a document actually contains readable text before designing another extraction workflow.
Use text positions and font identifiers to inspect how words are stored on a page.
Use the exported page records as a reviewed input to your own script, with source traceability.
PDF parsing and JSON assembly run in the browser. The PDF, metadata and extracted text are not uploaded. The downloaded JSON deliberately includes local source-file information.
Select a question to view the answer.
No. It focuses on pages, extracted text and optional metadata and text-item details.
No. It uses the text already present.
Yes. Disable text positions before conversion.
Yes, in the local report’s source object. It is not sent as a processing upload.
Open another Fylvexa tool that works well with this task.