Extract Text from PDF with Arabic Support

What does the extract text tool do?

A PDF written on a computer carries a text layer inside it: real characters, each with a position on the page. This tool reads that layer and rebuilds its lines in reading order, handing you text you can paste into a message, a document or a spreadsheet cell instead of retyping a contract or a report line by line.

All of the reading happens inside your browser, so the file is never uploaded and its text never passes through any server. That is what makes the tool usable on a contract, a bank statement or a medical report you do not want leaving your device. When it finishes, the text appears in a box you can read before you take it, with a copy button and a download button beside it.

When to extract text, and when to convert to Word or Excel

Extract text when you want the words and the shape of the page does not matter: a clause you are moving into an email, a term you are searching for, names and numbers going into a form, or text you would rather read in one block than page through. What you get is raw text, fast and light, with no fonts, no tables and no pictures.

Convert to Word when you want a document you will carry on editing as paragraphs, and to Excel when the page is a table whose columns need to stay columns. The simple rule: if the layout is part of what you need, this is not the tool, and if the words alone are the point, this is the shortest way to them.

If the document is a scan or a camera photo there is no text layer in it at all and nothing here to read. That is the job of OCR, and the tool says so plainly rather than handing back an empty file, with a link into the recognition tool that carries your document along.

Why Arabic comes out intact here

A shaped Arabic PDF stores its letters in visual rather than logical order, which is why Arabic falls out of so many tools reversed or broken apart. We assemble each line first as it sits on the page, then reorder an Arabic line right to left, and leave the numbers and Latin words inside it as islands in the order they were written.

That distinction is not cosmetic. A line reading that the total is 1,250.00 riyals used to come back with the amount reversed whenever the whole line was flipped, and a wrong figure in a contract is worse than an error message, because it looks perfectly plausible and so nothing tells anyone to check it. Figures now come out exactly as the author typed them.

On top of that we repair the lam-alef ligature: one glyph standing for two letters, transposed by any reader that reverses the line, which is how a correct word turns into a broken one. We compare the output against the glyphs on the page itself, fix that single case, and then normalize letter shapes back to their standard forms so search and paste behave in any program.

How to extract text from a PDF, step by step

Open the extract text page and drop a PDF into the tray, or tap it to pick the file from your device or from the Files app on a phone.
Type a page range if you only want part of the document, such as 1-3,5, and leave it empty to take every page. The box also accepts Arabic-Indic digits and the Arabic comma exactly as an Arabic keyboard produces them.
Drop additional files in alongside it if you want them read in the same run, and each file's name will appear above its own section of the output.
Press extract and wait a few seconds. The progress bar tracks the whole run rather than the page being read at that moment.
Read the text in the box that appears. Each line is shown in its own direction, which is what a document that mixes Arabic and English actually needs.
Press copy text to put it on your clipboard, or download it as a text file that opens with Arabic intact in Windows Notepad and in Excel.

Limits worth knowing

A scanned file has no text in it and this tool does not read pictures. It will not return an empty file: it tells you so and opens the way into OCR carrying your document with it.
Page layout is not preserved. A table arrives as a run of lines, two columns may interleave, and footnotes mix into the body. If the page is a table, the PDF to Excel tool is the right road, and if the layout matters, PDF to Word is.
A date that appears inside an Arabic paragraph in a different order from the way it was typed comes out as the page presents it, because that is genuinely what the page carries. Check dates before moving them into an official document.
Images, signatures and stamps are not extracted, because they are not text. To pull the pictures themselves out of a file, use the extract images tool.
A password-protected file will not give up its text layer until the password is removed with the unlock PDF tool, using the password you already know.
You can drop more than one file into a single run, and a run takes up to 3 files. The ceiling is technical rather than commercial, because every file is opened and read inside your own browser.
The summarize with AI button, where it appears under the result, is the one step that sends anything: the text alone and never the file, after you choose it and agree, every time. The extraction itself already happened on your device.

Common problems and their fixes

The text came back empty, or a message says the file has no text

The file is a scan or a set of images. Press the link inside that message: it carries your file into the OCR tool, and you pull the text out from there instead.

Numbers and amounts come out reversed

That does not happen here: the Arabic line is reordered while the figures inside it stay as the page shows them. If a number still looks reversed, compare it against the original page first, because a date may genuinely be laid out that way in the document itself.

My table arrived as lines and the columns are gone

That is what raw text does. Drop the same file into the PDF to Excel tool, which infers the columns from the whitespace on the page and puts every value back in its own cell.

The downloaded text file opens as symbols instead of Arabic

The file we produce carries a Unicode byte order mark at the front, which is what makes Windows Notepad and Excel read Arabic correctly. If you opened it in an older program that forces another encoding, open it in a Unicode-aware editor, or use the copy button and paste directly.

A message says the file is protected

The text layer is locked. Press the link in the message to reach the unlock tool with your file already in hand, type the password you know, then bring the result back here.

I want a Word document, not raw text

Do not copy from here and paste over there. Drop the file straight into the PDF to Word tool, which builds a docx with paragraphs out of the very same layer this tool reads.

The file is long and my phone slows down

Read it in batches with the page range: type 1-20, then 21-40, and so on. Every page is read inside your browser, and a file of several hundred pages is heavier on an older phone than a short one.

Frequently asked questions

Why is my file's text broken or empty?

If the file is a scan it holds images rather than text, and this tool reads the text layer, not pictures. It will not hand you an empty file: it says the document has no text layer and gives you a link into the free OCR tool that carries your file with it, so you continue in one tap instead of finding the file all over again.

Does Arabic come out in the right order?

Yes. Arabic lines are assembled right to left, while numbers and Latin words inside them keep the order the page shows. That is why an amount such as 1,250.00 comes out exactly as written rather than reversed, which is what happens when a reader flips the whole line and quietly corrupts every figure in a contract or an invoice.

Can I extract from specific pages only?

Yes. Type a page range in the box before extracting, such as 1-3,5, or 4- to take everything from the fourth page to the end. Leave the box empty and you get the whole document. The box also accepts Arabic-Indic digits and the Arabic comma, so an Arabic keyboard needs no switching for one field.

Why do other tools turn Arabic letters into shapes like a broken alef-lam?

A shaped Arabic PDF stores its glyphs in visual order, and a lam-alef ligature is one glyph standing for two letters, so a reader that reverses the line transposes them. We compare what came back against the glyphs on the page itself, repair only that case, and leave every other letter exactly as it was so real words are never rewritten.

Can I extract text from more than one file at once?

Yes. Drop several files in the tray and each is read in turn, then you get one block of text with each file's name printed above its own section so you can tell where one ends and the next begins. The number of files per run is capped for a technical reason: every file is opened and read inside your own browser.

How do I download text that opens in Arabic without strange symbols?

Press download and you get a Unicode text file with a byte order mark at the front. That mark is what makes Windows Notepad and Excel show Arabic properly instead of rows of symbols, and it is the single most common way Arabic text gets mangled on its way from one program to another through a plain text file.

The file is password protected, so what now?

The text layer will not open until the password is removed. The tool tells you the file is protected and offers a link into the unlock tool that carries the file along, so you type the password you already know there and come back with an open file. All of it happens inside your browser and the password never leaves your device.

Extract text from a PDF

Three steps, that's it