Pull the same fields out of a folder of PDFs, into one table
Invoices, statements, forms, applications: the fields you name are read from every file, and one spreadsheet comes out with a row per document.
Free: 3 jobs per UTC day, 25 MB per file. PDF exports include a visible Made with Bindery watermark. Pro exports clean PDFs.
Prefer the full studio? Open Extract fields to CSV in the app →
A folder of same-shaped documents — invoices, statements, application forms, exam papers — ends its life as a spreadsheet someone types by hand, opening each PDF and copying values across. Extract fields does the typing on your device: name the fields once, one per line, drop the folder, and every file gets a row. A field whose name is not what the page prints takes a label: “Sender | From” reads the line that prints From. Date-like fields read a date wherever a label is not found, and money fields take the largest amount on the document — the total, not the tax line.
Two engines read the pages. Text rules match labels on the text layer — deterministic, instant, no model, and it shows its working in the row order you can check against the files. The AI route reuses the engine behind Ask this PDF: the on-device model (a one-time download, nothing leaves the tab) or your own key, with the page text going only where you choose. Scanned pages are read with OCR inside the same run, while the daily OCR allowance covers them. The result is one ZIP with extracted-fields.xlsx and extracted-fields.csv — one row per file, one column per field, the file name first. Nothing is uploaded, and the whole batch counts as one of your daily jobs.
How it works
- Stage the PDFs, then open Extract fields → CSV and name the fields — one per line, an optional page label after a |
- Choose text rules or the AI engine and press Extract to a table
- Download the ZIP: extracted-fields.xlsx and extracted-fields.csv, one row per file
Questions people ask
Are my documents uploaded?
No. Reading, matching, OCR and the spreadsheet all happen in your browser tab. The AI route with the on-device model also runs entirely on your device; only if you choose your own API key does the page text go to the provider you chose, exactly as Ask this PDF says before you start.
What if a field is missing from one file?
Its cell stays empty rather than being guessed, so the gaps are visible in the table. Text rules read the value only when the label (or the field name) appears on the page; dates and totals fall back to the document’s date and its largest amount.
How does it count against the free plan?
The whole batch is one of your 3 daily jobs, files up to 25 MB, and up to 5 files per run on the free plan — Pro removes the cap. Scanned pages are OCR’d into the run and share the 5 free OCR pages a day with the OCR and Translate tools.