# PDF to Word: why the layout breaks, text PDF against scanned PDF, and how to get editable text out

> Why can you not just edit a PDF like a Word file? What a PDF actually stores, how to tell a text PDF from a scan in ten seconds, why Hindi PDFs sometimes paste as gibberish, and a practical way to get the text into Word and tidy it up. Honest about what plain-text extraction can and cannot do.

Published 2026-10-22 by Toolcloak. 4 minute read. Original: https://toolcloak.com/articles/pdf-to-word-why-the-layout-breaks-text-pdf-against-scanned-pdf-and-how-to-get-editable-text-out
Related free tool: [PDF to Word](https://toolcloak.com/tools/pdf-to-word)

A PDF is built to look the same everywhere. That is its strength and also why it is hard to edit. Inside the file there are no neat paragraphs, just instructions such as “draw this letter at this spot”. A converter has to **guess** the reading order, and the guess is only as good as the file allows. This guide explains what happens, how to tell what kind of PDF you have, and how to get usable text out.

> **In short**
> - **A text PDF** (saved from Word, a website or an app) lets you extract the words. **A scanned PDF** is only pictures, so you need OCR.
> - **Plain-text extraction gives you the words, not the formatting.** Bold, fonts, columns and tables are rebuilt by you in Word afterwards.
> - **Indian-language PDFs may paste as nonsense** if they use an old font. OCR with the right language fixes it.
> - Do the work in a browser tool so a private PDF never gets uploaded.

## Why a PDF is not “just a Word file”

**What each format stores**

| Word document | PDF |
| --- | --- |
| Structure | Positions |
| Paragraphs, headings, lists | Each line or letter at an exact spot |
| Text flows to fit the page | Looks identical on every device |
| Easy to edit | No real paragraphs to edit |

*To turn a PDF into editable text, software has to work out which lines belong together, which column comes first and what is a header.*

## Which kind of PDF do you have?

This decides everything. Open the PDF and try to **select a few words with the cursor** or search for one with Ctrl+F. If you can, it is a text PDF. If you cannot, or the cursor selects a whole picture, it is a scan.

**Four kinds of PDF and what to do**

| Box | What it means |
| --- | --- |
| Text PDF | Made from Word, Google Docs, a website or an invoice system. Words can be selected. Extract the text directly. |
| Scanned PDF | Photos or scans of paper. No selectable words. Use [Image to Text (OCR)](/tools/image-to-text-ocr) after turning pages into images. |
| Mixed PDF | Some pages typed, some scanned. Extract the text pages, OCR the rest. |
| Locked PDF | Needs a password to open or blocks copying. Unlock it first if you own it. |

*See the guides on [removing a PDF password](https://toolcloak.com/articles/how-to-remove-a-password-from-a-pdf-when-you-know-it-and-what-to-do-when-you-do-not) and on [OCR for English and Hindi](https://toolcloak.com/articles/how-to-copy-text-from-an-image-or-screenshot-ocr-in-english-hindi-and-other-languages).*

## Get the text out, step by step

**From PDF to a tidy Word document**

1. **Extract the text.** Open the PDF in the tool. All the text is pulled out in order.
2. **Check the result.** Scan for jumbled columns, repeated headers and broken lines.
3. **Copy into Word or Docs.** Paste and delete the repeated page headers and footers.
4. **Rebuild the formatting.** Apply headings, bold, bullets and tables. This is much faster than retyping.

*For a long document, do it in sections. Extracting one chapter at a time makes the check easier.*

Try it: [PDF to Word](https://toolcloak.com/tools/pdf-to-word). Extract all the text from a PDF into editable plain text, in your browser. Nothing is uploaded.

## What comes through and what does not

| In the PDF | In plain-text extraction | What to do |
| --- | --- | --- |
| Words and numbers | Yes, in reading order when the file is well built | Proofread once |
| Headings and bold | Come out as ordinary lines | Reapply styles in Word |
| Two-column pages | May be read straight across the columns | Copy each column separately if needed |
| Tables | Cells become lines of text | Rebuild the table, or try the PDF to Excel tool |
| Images and charts | Not included | Save pictures separately |
| Page numbers, headers | Repeat on every page | Delete them in Word with find and replace |
| Hyphenated line breaks | Line ends stay as breaks | Use find and replace to remove stray breaks |

*A tool that produces plain text is honest about this. A formatted .docx conversion tries to rebuild the layout and often makes a mess of complex pages.*

## Why Hindi text sometimes pastes as gibberish

Many older Indian documents use legacy fonts such as Kruti Dev or Chanakya. These fonts draw Devanagari shapes using the keys of an English keyboard. The PDF looks right, but the text inside is stored as Latin letters and symbols, so copying it gives something like “dsUnzh;” instead of केंद्रीय.

**Fixing legacy-font text**

1. **Spot the symptom.** The page looks fine, but pasted text is strange Latin letters.
2. **Treat it as a scan.** Turn the pages into images with [PDF to Images](https://toolcloak.com/articles/how-to-convert-a-pdf-to-jpg-or-png-pick-the-quality-choose-pages-and-keep-the-text-sharp).
3. **Run OCR with Hindi.** Choose English + Hindi. OCR reads the shapes, not the broken text layer.
4. **Proofread.** Check conjuncts and matras, which OCR can miss.

*Modern PDFs that use Unicode fonts such as Mangal or Noto Sans Devanagari copy correctly with no OCR.*

## Scanned PDFs: OCR is a different job

A scan is a photograph of a page. There are no letters in it, only pixels, so nothing can be extracted. OCR (optical character recognition) looks at the shapes and makes text from them. It works best on **straight, sharp, high-contrast scans at 200 to 300 dpi,** and badly on tilted, shadowed or handwritten pages.

- Scan flat, in good light, and keep the page square to the camera.
- Use black text on a white background where you can.
- Check numbers carefully after OCR. 1, l, 0 and O are easy to confuse.
- If you only need a few pages, [split them out first](https://toolcloak.com/articles/how-to-split-a-pdf-into-single-pages-or-equal-parts-and-pull-out-only-the-pages-you-need).

## Privacy and size

Contracts, bank statements and ID copies should not be uploaded to a random website. A browser-based tool reads the file on your own device and shows you the text. If the PDF is very large, [shrink it first](https://toolcloak.com/articles/how-to-compress-a-pdf-why-files-are-so-big-and-how-to-get-under-an-upload-limit) so it opens quickly on a phone.

> **A fast quality check:** After extraction, use Find in your editor for a rare word from the page. If it is found, the text is real. Also look at the first and last lines of two or three pages. Order problems usually show up there.

See also: [an ATS-friendly resume format](https://toolcloak.com/articles/ats-friendly-resume-format-for-freshers-and-experienced-what-to-write-what-to-avoid-and-how-to-save-the-pdf).

Related guides: [how to extract pages from a PDF](https://toolcloak.com/articles/how-to-extract-pages-from-a-pdf-keep-only-the-pages-you-need-in-any-order-and-save-them-as-a-new-file).

See also: [fixing the broken lines in text copied out of a PDF](https://toolcloak.com/articles/how-to-remove-line-breaks-from-text-copied-from-a-pdf-or-email-join-lines-into-one-paragraph-or-a-comma-separated-list).

## Frequently asked questions

### How do I convert a PDF to Word for free?

Open the PDF in a free browser tool that extracts text, copy the result into Word or Google Docs, then rebuild headings, bullets and tables. This works best on PDFs that were created from a document rather than scanned.

### Why can't I copy text from my PDF?

Either the PDF is a scan made of pictures, in which case you need OCR, or it is protected. If you own a protected file, remove the password first.

### Why does the formatting change when I convert PDF to Word?

A PDF stores exact positions, not paragraphs and styles. A converter has to guess the structure, so columns, tables and fonts often change.

### Why does Hindi text from a PDF look like random characters?

The PDF probably uses a legacy font such as Kruti Dev, which stores Devanagari as English-keyboard characters. Convert the pages to images and run Hindi OCR to get proper Unicode text.

### How can I tell if a PDF is scanned?

Try selecting a word or searching with Ctrl+F. If you cannot select text, it is a scan. Scans are also often much larger per page than typed PDFs.

### Is it safe to convert a PDF online?

It is safest with a tool that runs in your browser and does not upload your file. Avoid uploading ID, bank or legal documents to unknown sites.

### Can I convert a scanned PDF to editable text?

Yes, with OCR. Turn the pages into clear images, run OCR in the right language and proofread the output, especially numbers and names.