Which marks does this remove?

It removes every Arabic combining diacritic: the short vowels fatha, damma and kasra, the tanwin (double-vowel nunation marks), sukun, shadda, the superscript dagger alef, and the Quranic annotation signs. Base letters and their connections are left completely intact.

Does it change or break the base letters?

No. Only combining diacritic code points are deleted. The consonantal skeleton — alef, baa, taa and so on — is untouched, so the result is the normal unvowelled Arabic spelling people read every day.

What is tatweel and why is it optional?

Tatweel (also called kashida) is the stretching character `ـ` used to elongate letters for justification or calligraphic effect. It is not a vowel mark, so removing it is offered as a separate toggle in case you want to keep the original spacing.

Why would I remove harakat?

Fully vocalised text is useful for learners and recitation, but most digital Arabic — search queries, file names, body copy, database keys — is written unvowelled. Stripping the marks normalises text for matching, indexing, or a cleaner reading layout.

Is the text uploaded anywhere?

No. All processing happens locally in your browser using a Unicode character filter. The text you paste never leaves your device.

What is the Arabic Harakat Remover?

Remove all Arabic diacritics (fatha, kasra, damma, sukun, shadda, tanwin, dagger alef and Quranic marks) to produce clean unvowelled Arabic script. Optionally strip tatweel too. Runs entirely in your browser, nothing uploaded. It runs free in your browser on Gera Tools, with nothing uploaded.

Arabic Harakat Remover

Name: Arabic Harakat Remover
Creator: Gera Tools
License: https://creativecommons.org/licenses/by/4.0/

Clean, unvowelled Arabic in one step

Vocalised Arabic packs short vowels and other marks above and below the letters. That is essential for teaching, recitation, and disambiguation, but it gets in the way when you need plain text for search, matching, or a tidy layout. This tool deletes every diacritic and hands back the bare consonantal spelling.

How it works

Arabic diacritics live in well-defined Unicode ranges, separate from the base letters. The tool runs your text through a character filter that matches those combining marks and removes them:

U+064B..U+065F  fatha, damma, kasra, tanwin, shadda, sukun, …
U+0610..U+061A  Quranic annotation signs
U+0670          superscript (dagger) alef
U+06D6..U+06ED  Quranic small high/low marks

Because only these combining code points are targeted, the base letters and their joining behaviour are preserved exactly. A separate toggle removes the tatweel character ـ, which is a justification glyph rather than a vowel mark.

When you need unvowelled Arabic

Full-text search. Search indexes for Arabic content almost always work on unvocalised text, because users do not type harakat in search queries. If your stored content is vocalised, strip the marks before indexing so that the index and the query are in the same form. An Arabic user searching for كتاب will find كِتَابٌ only if the stored form has been diacritic-stripped.

Database normalisation and deduplication. Records that represent the same person, place, or product may arrive with different levels of vocalisation depending on the source. Stripping harakat and then comparing the results is the first step in collapsing near-duplicates.

OCR post-processing. Arabic OCR, particularly from older printed books or photocopies of typed documents, often produces inconsistent or malformed harakat marks. Stripping them gives a clean starting point for NLP work.

Copy-pasting into systems with poor Arabic support. Some CRMs, spreadsheets, and database clients misrender or misstore combining marks, causing text corruption. Removing the marks before pasting eliminates that class of problem.

Tatweel: why it is a separate option

Tatweel (ـ, U+0640) is not a vowel mark and not a combining character — it is a base character that stretches horizontal connections between letters, used in calligraphic typography and occasionally in headings. Because it does affect the text’s appearance rather than its phonology, some workflows want to keep it (for layout fidelity) and others want to remove it (to normalise spellings for matching). The separate toggle lets you choose.

Marks that are removed and preserved

The tool removes only combining marks — code points with zero-width rendering that sit above or below a base letter. Every base Arabic letter (alef, baa, taa, and so on), every letter’s connecting joins, and all punctuation survive unchanged.

The removal counter shown after processing gives you a quick sense of how heavily pointed the source text was. A high count from a short passage (for example 80 marks across 50 words) indicates fully pointed text such as a Quranic excerpt. A count of zero means the text was already unvocalised.

What this tool does not do

Removing harakat does not normalise alef variants (أ إ آ ا), standardise ta-marbuta, or fold alef-maksura. For that level of normalisation, pass the diacritic-free output through the Arabic Alef Normalizer tool as a second step.