Question 1

My table has visible borders — does that help?

Accepted Answer

Yes, a lot. When a page draws real rule lines — most invoices, bank statements and price lists do — the converter reads the exact row and column grid from those borders, cell for cell, instead of inferring columns from text alignment. Pages without ruling automatically fall back to position clustering, and you can turn ruled reading off from the options.

Question 2

Which PDFs convert cleanly?

Accepted Answer

The detection is rule-based geometry, not AI: it works best on clearly tabular PDFs — invoices, statements, price lists, reports with aligned columns. Pages that are plain prose become a single-column sheet (one row per line), and tangled or overlapping layouts may need a quick tidy-up after export.

Question 3

What happens to text that wraps inside a cell?

Accepted Answer

Multi-line wrapped text joins into one cell: a tightly-spaced continuation line whose columns sit under the row above merges into those cells ("Quarterly totals are audited" + "by an external accountant" becomes one cell). Fully wrapped rows where every column breaks at once are rarer and may land as their own row — a quick check after export catches them.

Question 4

Does it work on scanned PDFs?

Accepted Answer

No — a scan has no selectable text, so the detector has nothing to position. Run the file through the PDF OCR tool first (English + Hindi, on-device), then convert that text-based PDF here. The page-1 probe warns you before you waste a run.

Question 5

Are colours, borders and formulas recovered?

Accepted Answer

No — this is text and numbers only, honestly. Cell fills, borders, merged-cell styling, colours, images and any formulas behind the PDF's values are not recovered; what you get is the detected text grid with pure numbers as numeric cells. Re-apply formatting in Excel or Sheets after export.

PDF to Excel

How it works

Frequently asked