Get statements and ledgers into a sheet
Convert a bank or payment platform's PDF statement into xlsx for totals and filtering.
Loading…
Guide
PDF to Excel turns statements, reports and lists — PDFs where “the data is arranged in tidy rows” — into an .xlsx. It does not detect ruled lines; it reads the coordinates of every text block: blocks whose vertical positions are close are grouped into a row, the horizontal gaps that no text covers become column boundaries, and each row's text is cut into cells at those boundaries. Every page becomes a worksheet, and purely numeric cells are converted to numbers.
Updated 2026-09-094 min read
PDF to Excel turns statements, reports and lists — PDFs where “the data is arranged in tidy rows” — into an .xlsx. It does not detect ruled lines; it reads the coordinates of every text block: blocks whose vertical positions are close are grouped into a row, the horizontal gaps that no text covers become column boundaries, and each row's text is cut into cells at those boundaries. Every page becomes a worksheet, and purely numeric cells are converted to numbers.
This is a Beta tool whose inference algorithm is still being tuned. What you can adjust is the set of “column boundaries”: when the automatic result is not good enough, add or delete boundary lines by hand and apply them to all pages. A scan has no text coordinates to read, so the page says so and offers a one-click hand-off to “PDF OCR”.
x=value labels in pt; “Preview page” switches between pages.<name>_tables.xlsx with sheets named “Page 1”, “Page 2” and so on, skipping pages without text.| Input | Output | Notes |
|---|---|---|
| A bank statement PDF with six evenly spaced columns | statement_tables.xlsx, one sheet per page, the amount column numeric |
“1,234.50” loses its thousands separator and becomes a number |
| Two columns whose gap is too small and get merged | Click “Add” at an x coordinate between them | The boundary is applied to all pages |
| A report whose title spans the full width | The title is skipped automatically and takes no part in inference | Text blocks wider than 60% of the page are ignored |
Convert a bank or payment platform's PDF statement into xlsx for totals and filtering.
Convert a supplier's quotation or stock list PDF into a table instead of retyping it.
For a list whose pages share a layout, adjust the column boundaries once and export every page in one go.
Column boundaries are inferred from the gaps between text, so when some cell's content is long enough to fill a gap the boundary is missing or shifted. Go to “Column boundary adjustment”, delete the wrong boundary and add one in the right place — the change applies to every page.
So that they can be summed in Excel directly, purely numeric cells (allowing thousands separators, a minus sign and a decimal point, up to 15 digits) are written as numbers; identity numbers, very long identifiers and anything containing letters stay text.
The current version always produces one sheet per page; after exporting, copy the sheets together inside Excel.
The PDF is parsed and the xlsx is generated locally in the browser, and the file content is never uploaded to any server; the data is released from memory when the page is closed.
Updated 2026-09-09
Infer table rows and columns from text positions and build an xlsx (one sheet per page); column breaks adjustable
Infers rows and columns from text coordinates; suits PDFs with tidy data (invoices, reports, lists)