How to Convert an Arabic PDF to Word Without Broken, Reversed or Garbled Text

You convert an Arabic PDF to Word and open the result to find one of these: letters that won't join (ا ل ك ت ا ب instead of الكتاب), words spelled backwards, strings of nonsense like «ÇáãÚáæãÇÊ», or nothing usable at all — a blank page or an image you can't edit.

If you work with Arabic documents — as a translator, researcher, lawyer, student or a business with partners in the Middle East — you've probably met this problem. Most advice boils down to "try another converter". That rarely helps, because the cause is inside the PDF itself. Once you know the cause, you know the fix. The same principles apply to Persian, Urdu and, to a large extent, Hebrew.

Step one: diagnose the file in 30 seconds

Before trying any fix, open the PDF in your browser or a PDF reader, then:

1. Try to highlight a sentence. Does the selection follow the words?
2. Copy it and paste into a plain text editor or a search box.
3. Look at what you get.

This tiny test is the most important step in this guide, because it tells you which fix will work and which will waste your time.

Why Arabic text breaks inside PDFs

A PDF is a description of a page, not a text document

PDF was designed to print a page identically on any device. Internally it stores instructions like "draw glyph #54 from font X at coordinates (120, 300)". What matters is how the page looks, not what the text means. A Word file, by contrast, stores the text as meaningful characters and lets the software decide how to draw them.

To turn a PDF into Word, a converter has to reconstruct the text from those drawing instructions. For English this is usually straightforward: each letter has essentially one shape and runs left to right. Arabic adds three complications. (If you're curious about PDF internals in general, the PDF Association is an excellent resource.)

Complication 1: Arabic letters change shape

The letter ʿayn has four forms: isolated ع, initial عـ, medial ـعـ and final ـع. Unicode has one base character for ʿayn, and rendering software chooses the correct form on screen. But Unicode also contains separate codes for each pre-shaped form, called Arabic Presentation Forms.

Some software — especially older tools — writes these pre-shaped forms into the PDF instead of base letters. The PDF looks perfect, but when you copy or convert it you get "frozen" glyphs that may appear disconnected in Word, render in odd fonts, and break search and spell-check.

Complication 2: text direction

Modern software stores text in logical order — the order you read it — and the viewer flips it for display. Some older software stores visual order: left-to-right as it appears on screen. Paste that into an app expecting logical order and every word comes out backwards.

Complication 3: fonts without a dictionary

A PDF font can carry a table (called ToUnicode) mapping each glyph to the character it represents. If that table is missing or wrong, there's no way to know glyph #54 is ʿayn. The result is random symbols — the classic «ÇáÚÑÈíÉ» is Arabic text decoded with the wrong character encoding, a phenomenon known as mojibake.

Fix 1: Arabic OCR (most reliable for scans and broken encodings)

Optical character recognition ignores the file's internal data and "reads" the page as an image, the way your eyes do. That's why it works even when the encoding is broken or missing.

1. Open OCR — extract text.
2. Choose the image or PDF.
3. Pick the language: Arabic only for pure Arabic, or Arabic + English for mixed documents. Choosing correctly improves accuracy.
4. Press Extract text. The first run downloads the language data; after that, processing happens on your device.
5. Copy the text into Word and set the paragraph direction to right-to-left.

Tips for better Arabic OCR:

• Aim for scans around 300 DPI. Lower resolutions lose dots and diacritics.
• Straight pages beat skewed ones by a wide margin. Use Rotate PDF or rescan.
• Clean printed fonts (Arial, Tahoma, Traditional Arabic) read well. Calligraphic styles like Diwani or Thuluth read poorly in almost every OCR engine.
• Coloured backgrounds and watermarks hurt accuracy; converting to grayscale can help.

We cover this in depth in Extract Arabic text from images and scanned PDFs.

Fix 2: open the file in Google Docs

Google Drive can turn PDFs into editable documents, applying OCR where needed, and its Arabic support is generally good.

1. Upload the PDF to Google Drive.
2. Right-click it → Open with → Google Docs.
3. You'll get a new document with an image of each page followed by the extracted text.
4. Download as Word: File → Download → Microsoft Word.

Downsides: most formatting and tables are lost, the file is uploaded to Google's servers (avoid for confidential material), and there are size and page limits for this conversion.

Fix 3: open the PDF directly in Microsoft Word

Word 2013 and later can open PDFs and convert them into editable documents:

1. In Word, File → Open, and choose the PDF.
2. Word warns that it will convert the file. Click OK.
3. Review the result.

This works well for healthy text PDFs, especially ones originally created in Word, and keeps a fair amount of formatting. It won't do better than its source on scans or broken encodings.

Fix 4: normalize disconnected letters (for technical users)

If text pastes with separated letters in the correct order, presentation forms are almost certainly the cause. Unicode defines an official way to convert these back into base letters: NFKC normalization.

In Python it's one line:

import unicodedata fixed = unicodedata.normalize("NFKC", text)

After normalization the letters are base characters again, so Word joins them correctly and search and spell-check work. Not a programmer? Fix 1 (OCR) achieves the same in practice.

Fix 5: reversed text

If letters within words come out backwards, the file stores text in visual order. Options:

• OCR is the simplest, most reliable fix — it reads the image and outputs logical order.
• For a few sentences, retyping is faster than any technical fix.
• Programmers can reverse characters within each line, but watch out for numbers and embedded English, which flip differently.

Fix 6: extract raw text and format it yourself

Sometimes the fastest route is to give up on "preserving formatting":

1. Use PDF to text to pull out just the text, if the PDF is healthy.
2. Or use PDF to Word (text only) for a DOCX containing the text.
3. Apply clean formatting in Word: one font, right-to-left direction, built-in heading styles.

You'll get a cleaner, more editable document than a "layout-preserving" conversion full of floating text boxes that jump around with every edit.

Why no converter preserves layout 100%

Many sites promise "perfect conversion with formatting intact". The truth is that a PDF doesn't record "this is a paragraph", "this is a heading" or "this is a three-column table". It only records where shapes sit on the page. Converters rebuild structure by guessing:

• Side-by-side columns can merge into interleaved text.
• Tables can become scattered text boxes.
• Spaces between Arabic words can vanish or double.
• Footers and page numbers can land in the middle of paragraphs.

With Arabic these problems multiply because of right-to-left direction mixed with left-to-right numbers and Latin words. Always plan for a manual review proportional to the document's importance.

Real-world scenarios

An Arabic contract from a partner, needing a few edits

It's usually a text PDF created in Word. Run the copy-paste test: if it passes, open the file in Word and edit. If the formatting is complex (tables, numbered clauses), it's often faster — and perfectly normal in negotiations — to ask for the original Word file.

An old scanned Arabic book you need to quote

OCR only the pages you need. Choose Arabic only, proofread quotes letter by letter (especially diacritics in literary text and poetry), and cite the page number from the original.

An official scanned letter you need to rewrite

OCR, then retype in your own Word template. Don't try to preserve the original look; letterheads and stamps don't convert to text.

A price list or transcript table

Tables are the hardest thing to convert. Extract the text and rebuild the table in Word or Excel. For large tables, crop each column into its own image, OCR it separately, and place the columns side by side.

A mixed Arabic–English technical report

Choose Arabic + English in OCR. After pasting into Word, check lines that start with an English word — Word may flip the paragraph direction. Select the paragraph and set right-to-left manually.

Quick fixes inside Word after converting

1. Paragraph direction: select all (Ctrl+A) and click "Right-to-left text direction" on the Home tab.
2. Alignment: right-align or justify for Arabic text.
3. Font: pick one Arabic font — Simplified Arabic, Traditional Arabic or Sakkal Majalla — and apply it to everything.
4. Extra line breaks: converters often put a hard break at the end of every original line, splitting one paragraph into ten. Use Find and Replace to swap paragraph marks (^p) for spaces inside paragraphs, then re-split real paragraphs by hand.
5. Double spaces: replace two spaces with one, and repeat until nothing is found.
6. Digits: if numbers display as Arabic-Indic (٠١٢) when you want Western (012), or vice versa, check the "Numeral" setting in Word's advanced options.

Why free converters struggle more with Arabic than English

Most PDF-to-Word engines were built for Latin scripts first, with Arabic support added later. The failure patterns repeat:

• Trusting the internal text: the converter copies whatever text is stored in the PDF — presentation forms and visual order included.
• Rebuilding layout left to right: it sorts text boxes by horizontal position, so column order flips in Arabic documents.
• Substituting non-Arabic fonts: if the original font isn't available, a fallback with weak Arabic support makes letters look disconnected.

That's why OCR sometimes beats a "smart" converter: it ignores faulty internal data and starts from what the eye sees.

Quick decision guide

Prevent the problem when you create Arabic PDFs

If you're the one producing Arabic PDFs, a few habits make them easy to convert later — for you and for others:

• Save as PDF from Word or Google Docs directly, not via "print to PDF" from older software.
• Use modern Arabic fonts with full Unicode support.
• Don't convert text to outlines or images in design tools unless you must.
• Keep the original Word file alongside the PDF. Converting PDF back to Word is a rescue operation, not a workflow.
• For long-term archiving, consider PDF/A, which requires embedded fonts and Unicode mapping.

Privacy: be careful where you upload documents

The Arabic PDFs we need to convert are often contracts, official letters or personal records. Most popular converters upload files to their servers. WPPR's OCR and PDF to Word tools process files in your browser. For OCR, only the language data is downloaded; your document stays on your device. See our security page.

FAQ

Why do Arabic letters appear disconnected after converting to Word?

Because the PDF stores pre-shaped presentation forms instead of base letters, or the font lacks a correct Unicode mapping. OCR is the simplest fix; programmers can apply NFKC normalization.

What causes gibberish like ÇáÚÑÈíÉ?

Arabic text read with the wrong encoding, or a font without a Unicode map. There's no reliable way to repair the extracted text itself — OCR the page instead.

Can a scanned Arabic PDF be converted into editable Word?

Yes, with Arabic OCR. Accuracy is excellent on clear print at 300 DPI and drops with handwriting and poor images.

What's the best software for converting Arabic PDF to Word?

There isn't one tool for every case. Opening healthy text PDFs in Word works well; for scans and broken encodings, nothing beats Arabic OCR. Choose based on the copy-paste test above.

Will tables survive the conversion?

Rarely intact with Arabic. Expect to rebuild tables in Word or Excel.

Does this apply to Persian, Urdu and Hebrew too?

Largely, yes. Persian and Urdu use the Arabic script with the same shaping and direction issues. Hebrew doesn't change letter shapes as much, but suffers the same visual-order and encoding problems.

Bottom line

Broken Arabic after conversion isn't a curse or a fault on your computer — it's a consequence of how text is stored inside the PDF. Always start with the copy-paste test: if nothing copies or you get gibberish, use OCR; if text copies cleanly, open it in Word or use a direct conversion; if letters are disconnected or reversed, normalization or OCR will fix them.

Start with OCR — extract text for scans, or PDF to Word for text PDFs — free, and nothing is uploaded.

Sources

• Unicode: Arabic Presentation Forms-B code chart
• Unicode Normalization Forms (UAX #15)
• Unicode Bidirectional Algorithm (UAX #9)
• Tesseract OCR: improving output quality
• Mojibake (Wikipedia)