What does this tool do?

It converts your PDF into an Excel spreadsheet (.xlsx). Each PDF page becomes a separate worksheet with rows and columns reconstructed from the page content.

Does the tool upload my PDF to a server?

No. Everything runs entirely in your browser. Your PDF is never uploaded, never stored, and never shared with anyone.

Will tables be reconstructed correctly?

Tables with clear column alignment usually convert cleanly. Numbers are detected automatically and stored as numeric cells. Heavily image-based PDFs or scanned pages without selectable text will not produce useful spreadsheet data.

Does each PDF page become a separate sheet?

Yes. Every page in the source PDF is placed on its own worksheet, named Page 1, Page 2, and so on.

How does the tool decide where columns are?

It looks at where text starts across the whole page. When many rows begin a value at roughly the same horizontal position, that position is treated as a column edge. A candidate edge has to appear in a meaningful share of the rows before it counts, which stops a single stray line from inventing a column. Where no consistent edges emerge, the tool falls back to splitting each row on its wider gaps.

Why did my table come out in one column?

The values were not vertically aligned enough to form an edge, which usually means the table is separated by ruled lines rather than by consistent spacing, or the columns are narrow and ragged. Landscape orientation and wider column spacing in the source both improve the result noticeably.

Are numbers usable in formulas?

Yes. Values recognised as numeric are written as real numbers rather than text, so sums and averages work immediately without a cleanup pass. Check currency and thousands separators once, since a symbol attached to a value can keep it as text.

What if the PDF is scanned?

You will get a workbook containing a note that no extractable text was found. A scan holds a picture of a table, not the values in it, and this tool does not perform optical character recognition.

Yes, completely free. No account, no sign-up, no watermark, and no row limit.

Document Tool

PDF to Excel Converter

Convert your PDF to an Excel spreadsheet (.xlsx) directly in your browser. Tables and rows are extracted into editable sheets. No uploads, no sign-up.

Back to all tools

PDF File

Choose filedrag & drop or Ctrl+V / ⌘V to pasteBrowse

How It Works

Browser-native table extraction

Your PDF is parsed inside your browser tab. Text is grouped into rows and columns based on its position on the page, then reassembled into a clean Excel workbook. One worksheet per PDF page.

Read The Workflow

Privacy

Your PDF never leaves your device

All processing happens entirely inside your browser tab. Your file is never uploaded, never stored, and never shared with anyone.

Review Privacy Policy

Getting a table out of a PDF and into real cells

A table in a PDF is an illusion. The file records a set of text fragments at fixed coordinates, and your eye assembles them into rows and columns. Nothing in the document says this is a table, which column a value belongs to, or where one record ends. Turning that back into a spreadsheet means rediscovering a structure that was thrown away when the PDF was created.

How the reconstruction works

Rows come from vertical position. Text fragments sharing a baseline are grouped into one row, then ordered left to right.
Columns come from repetition. The tool collects the horizontal start position of every fragment on the page and looks for positions that recur across many rows. Those recurring positions become column edges, so a genuine column has to prove itself across the page rather than appear once.
Gaps are the fallback. When no consistent edges emerge, each row is split on its wider internal spaces instead. Less precise, but it keeps loosely structured pages usable.
Numbers are typed, not just copied. Values that read as numeric are written as numbers, so totals and averages work without a cleanup pass first.
One page, one worksheet. Each PDF page becomes its own sheet, named Page 1, Page 2 and so on, which keeps a multi page report navigable.

What converts well, and what fights back

Source	Expected result
Bank statements, trial balances, exported ledgers	Strong. Machine generated tables have rigidly aligned columns, which is exactly what the detection looks for.
Invoice line items	Good for the line table. Surrounding address blocks and totals may land as loose rows around it.
Tables ruled with lines but loosely spaced	Mixed. The rules are graphics, not data, so alignment rather than the visible grid decides the outcome.
Merged cells and stacked headers	Flattened. A two row header usually becomes two ordinary rows to tidy up.
Values wrapped onto two lines in one cell	Split across rows, because the second line looks like a new record from the outside.
Scanned pages	Nothing usable. There is no text to read and no optical character recognition here.

Improving the odds

When you control how the PDF is produced, small choices upstream pay off. Exporting a wide table in landscape keeps columns apart instead of squeezing them until their edges blur. Turning off cell wrapping keeps one record on one line. Leaving generous column spacing gives the edge detection something unambiguous to find.

When you do not control the source, cropping helps. Removing headers, footers and margin notes with Crop PDF takes stray text out of the page before conversion, so it stops competing with the real columns. On a long report, isolate the pages that actually hold the table with Split PDF first.

Check before you calculate

Any extracted table deserves one verification pass before it feeds a decision. Compare the row count against the source, add up a column you already know the total of, and look at the first and last record on every sheet. Reconstruction errors cluster at page boundaries and in the widest columns, so those are the places to spend your attention.

Frequently Asked Questions

What does this tool do?: It converts your PDF into an Excel spreadsheet (.xlsx). Each PDF page becomes a separate worksheet with rows and columns reconstructed from the page content.
Does the tool upload my PDF to a server?: No. Everything runs entirely in your browser. Your PDF is never uploaded, never stored, and never shared with anyone.
Will tables be reconstructed correctly?: Tables with clear column alignment usually convert cleanly. Numbers are detected automatically and stored as numeric cells. Heavily image-based PDFs or scanned pages without selectable text will not produce useful spreadsheet data.
Does each PDF page become a separate sheet?: Yes. Every page in the source PDF is placed on its own worksheet, named Page 1, Page 2, and so on.
How does the tool decide where columns are?: It looks at where text starts across the whole page. When many rows begin a value at roughly the same horizontal position, that position is treated as a column edge. A candidate edge has to appear in a meaningful share of the rows before it counts, which stops a single stray line from inventing a column. Where no consistent edges emerge, the tool falls back to splitting each row on its wider gaps.
Why did my table come out in one column?: The values were not vertically aligned enough to form an edge, which usually means the table is separated by ruled lines rather than by consistent spacing, or the columns are narrow and ragged. Landscape orientation and wider column spacing in the source both improve the result noticeably.
Are numbers usable in formulas?: Yes. Values recognised as numeric are written as real numbers rather than text, so sums and averages work immediately without a cleanup pass. Check currency and thousands separators once, since a symbol attached to a value can keep it as text.
What if the PDF is scanned?: You will get a workbook containing a note that no extractable text was found. A scan holds a picture of a table, not the values in it, and this tool does not perform optical character recognition.
Is this tool free?: Yes, completely free. No account, no sign-up, no watermark, and no row limit.