Utilore

PDF to Text

Copying from a PDF gives you words cut in half, lines broken at the column edge and a company name repeating every forty lines. Here the text is pulled out and cleaned up — each fix is a switch you turn on yourself, and the file never leaves this tab.

or drop them here

A PDF with a text layer — not a scan. Up to 60 MB

The file stays on this device: it is opened inside the tab and nothing is uploaded. Nothing is downloaded either — the PDF reader is already part of this page.

What does not carry over: pictures, tables as tables, bold and italic. A table becomes rows of text, and a two-column layout is read left column first — the way the text is stored in the file.

Nearby: PDF to Word · Image to Text · Compare Two PDFs · Word Counter

How to use

  1. Choose a PDF that has a text layer — not a scan.
  2. Turn on what you need: join hyphenated words, join lines into paragraphs, drop headers and page numbers.
  3. Copy the text or download it as a .txt file.

Good to know

Getting the text is half the work

A PDF does not store lines — it stores pieces of text with coordinates. One line often arrives as a dozen pieces because the font changed mid-sentence, and a naive extractor returns them one per line. So the pieces are first grouped back into lines by their vertical position, and only then cleaned. The same grouping runs in the tool that compares two PDFs: if it worked differently there, the same file would show different differences.

Three fixes, and none of them happens silently

Hyphenated words are joined — but only when the next line starts with a lowercase letter, so a compound name is not glued into one word. Lines are joined into paragraphs — but headings, numbered clauses and list items keep their own lines, because merging the clauses of a contract is worse than leaving extra line breaks. Headers and page numbers are dropped by repetition across pages — and a bare number counts as a page number only if it is not larger than the page count, so the year 2026 in a table survives.

A scan gets a different answer

If the PDF is a scan, there is no text in it at all — the pages are pictures. Returning an empty box would look like a broken tool rather than the wrong tool, so the page says what happened and points to text recognition, which reads PDF pages as well. That check is deliberately blunt: fewer than forty letters in the whole document means captions inside images, not a text layer.

Frequently asked questions

Is my document uploaded anywhere?

No. It is opened inside this browser tab and read there. Nothing is downloaded either — the PDF reader is already part of this page.

Why is the text empty for my file?

Most likely it is a scan: a picture of each page with no text layer. Run it through “Image to text” instead — that tool recognises the letters, and it takes PDF pages too.

Why are the columns mixed up?

Because a two-column layout is stored in the file the way it was drawn, and the left column usually comes first as a whole. The order here follows the file rather than guessing at the layout — guessing wrong scrambles the text more than the columns do.

What happens to tables?

A table becomes rows of text: the cells keep their order but lose their grid. If you need the grid, use “PDF to Excel” — it works out the columns from the coordinates.

How large a document can it take?

Up to 60 MB and the first 500 pages. Beyond that the wait stops being worth it, and the page says so instead of freezing.

Related tools