PDF to Excel Converter – Extract Tables from a PDF into XLSX

Extract a table from a PDF into a spreadsheet

This tool takes a PDF that contains rows and columns of text and writes them into an Excel workbook. You add the file, press one button and download an .xlsx with one sheet for every page that has text. It is the quickest way to convert PDF data to a spreadsheet when the numbers were only ever given to you as a PDF.

The people who need it are not hard to guess: an accountant handed a bank statement or a GST summary as a PDF, a teacher with a marks list, a student with a timetable or a result sheet, a shop owner with a supplier price list, an office clerk with a list of names and addresses. In each case the data is already laid out in a grid, and retyping it is the slow part.

How to turn a PDF table into an Excel file

1. 1
   Drop the PDF on the box that says Drop files here, or click to browse, or click the box and choose it. You can add more than one PDF.
2. 2
   Read the line under the box. It reports the pages and whether the file has selectable text. For a scan it warns that there is no table to pull out.
3. 3
   Press Convert to Excel. The button shows Working with a percentage as each page is read.
4. 4
   Read the note in the result panel. It tells you how many sheets were made and whether a table structure was detected.
5. 5
   Use Download to save the .xlsx, which takes the PDF's name. Several PDFs can be saved together with Download all as a ZIP.

How rows and cells are worked out

The PDF is opened with pdf.js and, for each page, the tool asks for every piece of text and its position. Pieces of text with the same baseline, rounded to a whole point, are put in one row from top to bottom of the page. Inside a row they are ordered left to right, trimmed, and each one is written to its own cell with the SheetJS library. Blank pieces are dropped and rows with nothing in them are skipped.

This is worth understanding because it explains everything you see. The tool does not draw a grid and snap text into it. It does not read ruling lines, and it does not know that a cell is empty. Two things that were far apart on the page become neighbouring cells, and a value that the PDF stored as one piece of text, such as a whole address, stays in one cell.

A worked example: a short bank statement

Take a page with a header row and three transactions, where each row has a Debit column and a Credit column and one of the two is always empty. The first row has four pieces of text. The rows below have only three, because the empty column contributes nothing. This is what came out of a real run of the converter.

Three things happened. The salary credit of 52,000 now sits in the Debit column. The footer became a data row. And every figure is stored as text, so summing the three amounts gives 0 rather than 69,450. The conversion has not failed; it did what it does. You now need about two minutes of cleanup, not twenty minutes of typing.

Fixing numbers that arrive as text

Every cell comes through as text, including figures. You can usually tell because they sit on the left of the cell and Excel may show a small green triangle. Select the column, then use Data, Text to Columns and press Finish without changing anything. This forces Excel to re-read the entries and turns most of them into numbers. If a figure stays on the left, the usual culprits are a currency symbol, a stray space or a comma in a position your regional setting does not accept. Find and Replace can remove the symbol or the commas.

Check the result with a quick sum. In the example above, once 52,000 is moved to the Credit column and the debits 15,000 and 2,450 are converted, the debit total should read 17,450. If the sum is zero, the cells are still text.

Reading the note under the result

The note is the converter's honest summary. If it says Text runs became cells, at least two rows on some page held two or more pieces of text, so the file looks like a table and well-aligned ones come out cleanly. If it says No table structure was detected, every row is a single piece of text and each line sits in column A. You still get a workbook, and it may be exactly what you need for a list of names, but for running paragraphs a Word file or Markdown text reads better.

The sheets are named Page 1, Page 2 and so on, matching the PDF. A page with no text produces no sheet, so a PDF whose second page is a blank divider gives sheets named Page 1 and Page 3. Do not assume the sheet number equals the number of sheets.

Tables that run across several pages

Each page is kept separate on purpose. The tool does not guess where one table ends and the next begins. For a statement that runs over five pages you will have five sheets, each with its own page header and footer lines. Copy the data rows from sheets two to five under the rows of sheet one, delete the repeated column headings, and remove the footer rows.

If only part of a long file matters, save yourself the effort by pulling out the pages that hold the table before converting. If the table spans a mix of files, merge them first only when they share the same layout.

Which PDFs convert well and which do not

• Good: statements, invoices, price lists, mark sheets and timetables that were exported from software. Their text is real, their rows share a baseline, and their columns are separated by gaps.
• Mixed: tables with merged headings, multi-line cells, or empty cells. The text is read, but cells may shift or split.
• Poor: paragraphs of prose. Each line becomes one cell in column A, which is rarely useful.
• Not possible: scans and photographs of tables. They contain no text, and running the tool ends with a message saying no text was found. Read the scan with OCR first, or retype the data.

A five-minute cleanup routine in Excel

• Keep the original PDF open on the other half of the screen and compare the first and last rows of each sheet.
• Delete the rows that are not data: page numbers, running headers, company names and totals that you plan to recalculate.
• Insert a column wherever a value has slid, and move the stray cell across. Do this before you sort or filter anything.
• Convert the number columns with Text to Columns, then check with a sum against the total printed on the PDF.
• Format the dates. A date such as 02-Apr is usually still plain text until you set a date format or re-enter it.
• Save the file as a normal workbook under a new name, keeping the raw conversion untouched.

Choosing a good source PDF

Results depend more on how the PDF was made than on anything you can set. A statement downloaded from a bank portal or exported from accounting software holds proper text in tidy rows, and it converts well. A PDF created by printing a spreadsheet to PDF is also reliable. A PDF made by photographing a printout and saving it is a picture, so it fails.

If you have a choice, ask for the original format. A CSV or XLSX from the source system will always beat a conversion. When only the PDF exists, download it again from the portal rather than using a screenshot, a phone scan or a PDF that has been through a print-to-image step. A quick test is to try to select a single number with your cursor. If it highlights on its own, the converter can read it.

Where people use it

• Bookkeeping: bank and card statements, vendor ledgers and GST or TDS summaries that arrive as PDFs.
• Schools and colleges: marks lists, attendance sheets, fee registers and exam seating plans.
• Shops and wholesalers: supplier price lists and stock sheets that need sorting or filtering.
• Offices: contact lists, visitor logs and shift rosters circulated as PDFs.
• Research: tables from reports or papers that you want to analyse, checked against the printed figures.

Common mistakes

• Trusting the columns without checking. Always compare the first and last row of every sheet with the PDF.
• Forgetting the repeated headers and footers, which end up as ordinary rows in the data.
• Sorting before cleaning. If cells have shifted, sorting by a column will scramble the records.
• Treating text numbers as numbers. Convert them before you total anything.
• Using a locked file. An encrypted PDF is blocked on the page, so unlock it first and convert the open copy.

What the converter cannot do

It does not read borders, shading or colours, so none of the table's styling reaches the spreadsheet. It does not merge cells. It does not recognise formulas, because a PDF holds only the printed values. It does not read text inside images, and it has no OCR. There are limits on size too: a file can be up to 100 MB and 500 pages, because the work is done by your browser.

If you want the words from the PDF as paragraphs rather than cells, convert it to Word instead.

Privacy

The PDF is read and the workbook is built on your own device. Statements and mark lists are never sent to a Jobsixo server, and nothing is stored after you close the tab.