PDF to Excel

How to turn a PDF table into Excel

1. Choose the PDF (or drop it on the box), such as a bank statement downloaded from your bank. Its table comes out at once, shown as it will be in Excel.
2. Choose One sheet (every page's rows in one table, the header once) or A sheet for each page. To take only some pages, choose Some pages and type them.
3. Download the Excel file (.xlsx) or a CSV, or share it.

Everything happens in your browser: the PDF is never uploaded to ap7am or anywhere else. Your statement and its password stay in your browser, and neither is saved. The file keeps your PDF's name, such as statement.xlsx.

Only the parts you need, such as Domestic Transactions

1. Under What to convert, choose Only the parts I select. The tables in your PDF are listed under Tables found, with their pages and rows; a table that goes on over several pages is listed once, wherever it sits on each page.
2. Tick the one you want, such as Domestic Transactions. Its boxes show on the pages, numbered, to check, move or resize. Nothing from the summary boxes, adverts or other tables comes with it.
3. For anything else, draw a box over it: drag with the mouse, or on a phone tap ✏︎ Draw a box first and drag with your finger (or tap a table to take all of it). A box drawn mostly over a table fits itself to the table's edges; Undo once keeps it as you drew it. Same place on next pages copies a box to the pages where its table goes on.
4. The sheet holds the parts' rows in the order of their numbers, with the header row once. Untick that to keep each part's own header, or choose A sheet for each part.

A box moves and resizes with the mouse or a finger, or from the keyboard: the arrow keys move it, Alt and the arrow keys resize it, Delete removes it, and Ctrl+Z undoes. × removes a box, and the list under the page jumps to each part. A word only partly inside a box is taken when most of it is inside; a word is never cut in half.

How it finds the table

• Columns from where the text lines up: the gaps that stay empty down the page are the gaps between columns, even where one long description runs into the next column. A column that is mostly empty, such as Cheque No, is still a column.
• Rows: a description that wraps onto a second or third line is joined into its row's cell, whether the extra lines are below the date or above it.
• The header row (Date, Narration, Debit, Credit, Balance and the like) is found and kept once; the same header on later pages is left out, and so are the bank's name and page numbers repeated on every page. Tick Keep the text above and below the table for the account details and notes.

Amounts and dates as numbers

Amounts become numbers you can add up, and dates real Excel dates you can sort:

Indian commas (1,23,456.78) and international ones are both read; ₹, Rs. and INR before an amount (also a ₹ drawn with a font of its own, as some banks' statements do, which other tools read as a letter) and /- after it are taken off; a minus before or after, or brackets, make it negative. Dr makes an amount negative and Cr positive, and the cell still shows Cr or Dr. Where credits are marked with a + before the amount and debits with nothing, a Cr/Dr column is added and the amounts signed the same way. A column of coloured dots alone is left out. Dates are read day first, as in India (01/04/2025 is 1 April), unless a column's dates can only be month first; a time after the date (30/09/2026| 11:38) is kept, as an Excel date and time. A card holder's name printed over their transactions stays a row of its own, in bold. Long account and reference numbers, and numbers starting with 0, stay as typed, so Excel does not round them.

What it cannot do

• Scanned statements: a scan is a picture with no text in it. The tool says which pages have none; run it through OCR PDF first: it reads the Telugu and English in pictures.
• Tables are found from how the text lines up, not from the lines drawn around them, so cells merged across columns come out in one of them. Check the totals against the PDF before you rely on them.
• A CSV holds one sheet: with a sheet for each page, the CSV has every page's rows one after another.

Questions

Made with Mozilla's pdf.js (Apache 2.0 licence), running in your browser.