How to remove Arabic diacritics and clean up Arabic text
The Cleaner gathers the fixes Arabic text usually needs after it has been copied from a website, a document or a chat app: removing diacritics (harakat), deleting tatweel, converting digits between ١٢٣ and 123, and tidying spaces and line breaks. Reach for it when you need unvowelled text for search, when an online form rejects a phone number typed with Arabic-Indic digits, or when a social post arrives full of doubled spaces and stretched words.
Nothing changes unless you ask for it. Your original stays untouched, and the result appears beside it with a summary of every change.
How to use it
- Paste your text into Original text, or click Open .txt to load a plain-text file from your device.
- For the common jobs, click one quick action: Diacritics off, Tidy spacing or Digits to 123. A quick action replaces your current options, and Undo brings them back.
- For finer control, choose options yourself under Transform options: Arabic marks for diacritics and tatweel, Convert digits, and Spacing & line breaks. The result updates with every click.
- Check the Change summary for a count of each type of change, and open Compare changes to see removed text crossed out and added text underlined.
- Click Copy result or Download .txt. If the result is not what you wanted, use Undo or Reset options.
Worked examples from the tool
قَالَ الْمُعَلِّمُ: اقْرَأْ بِتَأَنٍّ.
قال المعلم: اقرأ بتأن.
Settings: Arabic marks · Remove common Arabic diacritics
رقم الطلب: ٤٥٨٢١ الجوال: ٠٥٥١٢٣٤٥٦٧ المبلغ: ١٢٫٥٠ د.ك
رقم الطلب: 45821 الجوال: 0551234567 المبلغ: 12٫50 د.ك
Settings: Quick action Digits to 123 · Western digits, links protected
عــروض نهاية الأسبوع خصم 30% على كل المنتجات
عروض نهاية الأسبوع خصم 30% على كل المنتجات
Settings: Remove tatweel · Trim spaces at line edges · Collapse spaces and tabs
What each option does
- Remove common Arabic diacritics deletes fatha, damma, kasra, tanween, shadda, sukun and the superscript (dagger) alef. Hamza and maddah stay, so أ إ آ ؤ ئ are untouched.
- Remove tatweel deletes the kashida stretch character ـ, so أهــلاً becomes أهلاً. It is separate from diacritics.
- Repair legacy Arabic presentation forms is for text copied from some PDFs that looks fine but cannot be found by search. If the words also run backwards, use Reversed Arabic instead.
- Convert digits maps every digit to the same value in the set you choose: Western, Arabic-Indic, or Extended Arabic-Indic (the digits used for Persian and Urdu). Tick Protect detected links & emails during digit conversion to keep the digits in URLs and email addresses as they are.
- Spacing: Trim spaces at line edges and Collapse spaces and tabs deal with extra spaces; Convert non-breaking spaces handles the special spaces that collapsing leaves alone. Under Line breaks, choose Keep at most one blank line or Join lines with a space.
More controls adds blank-line and duplicate-line removal, line sorting, and removing the space before Arabic commas, semicolons and question marks. A literal Find & replace panel sits below the editor. For a fuller discussion of when to strip harakat and when to keep them, read removing Arabic diacritics.
Common pitfalls
- Normalizing letters in text you will publish. The Normalize letters for search options change spelling: with them on, مدرسة becomes مدرسه and إلى becomes الي. Use them only on a copy meant for search or matching, as the search normalization guide explains.
- Joining lines in structured text. Addresses, poems and subtitle files depend on their line breaks, and joining turns them into one long line.
- Expecting the decimal separator to change. Digit conversion leaves ٫ alone. If a system rejects 12٫50, open Find & replace and replace ٫ with a full stop.
- Stripping vowels that carry meaning. Without harakat, علم can be read several ways. Keep a vowelled copy of religious texts, poetry and children's books.
For the full routine, see cleaning up copied Arabic text.
Limits
- Each text can be up to 100,000 text units. A unit is usually one letter, mark or space; an emoji can take two or more.
- It does not add diacritics, and it does not correct spelling or grammar.
- It cannot rebuild missing letters or gibberish copied from a PDF, and it never reverses text.
- Common diacritic removal does not cover every Quranic annotation mark, such as pause signs and small high letters.
- Line sorting follows your browser's Arabic ordering rules, which can differ slightly between browsers.
- Nothing is saved automatically. Closing or reloading the page clears your text unless you download it or save a workspace file.
Frequently asked questions
Does removing diacritics delete hamzas?
No. Remove common Arabic diacritics deletes only short vowels, tanween, shadda, sukun and the dagger alef. Hamza on alef, waw and yeh stays, and so does the maddah in آ.
How do I convert Arabic numbers like ١٢٣ to 123?
Click the Digits to 123 quick action, or under Convert digits choose Western — 0123456789. It converts Arabic-Indic and Persian digits in one pass. To go the other way, choose Arabic-Indic — ٠١٢٣٤٥٦٧٨٩.
I collapsed spaces but some gaps remain. Why?
They are probably non-breaking spaces, which can come from web pages and word processors. Tick Convert non-breaking spaces together with Collapse spaces and tabs. To find any other hidden gap, use the invisible character inspector.
Is my text uploaded anywhere?
No. Processing happens in your browser, and the app does not send your text or file to any server. See the privacy page.
Can I clean several files at once?
Yes. Set your options here, then open Batch files and choose Current cleaner settings. It accepts up to 20 .txt files and gives you the results as a ZIP.
Can I undo after several changes?
Yes. Undo keeps up to 50 steps during the session. Reset options returns every option to its default without touching your original text.