Tashkeel, also called harakat, are the small diacritical marks written above and below Arabic letters to indicate short vowels, gemination (shadda), and the absence of a vowel (sukun). Most everyday Arabic omits them and relies on the reader to supply vowels from context.

Why does diacritic density matter?

Fully vocalised text is far easier for learners and is required for the Quran, children's books, and ambiguous words. Native adult readers usually read unvocalised text fluently, so density is a strong signal of the intended audience.

Which marks count as diacritics here?

The meter counts fatha, kasra, damma, their tanwin forms, sukun, shadda, the superscript alef, and related Unicode marks in the U+064B–U+0652 and related ranges. A word counts as vocalised if it carries at least one such mark.

Does it change my text?

No. The tool only reads and counts marks; it never adds, removes, or rewrites diacritics. Everything runs locally in your browser and nothing is uploaded.

What is the Arabic Diacritic Density Meter?

Measures the proportion of vowel-marked (tashkeel) words to unvowelled words in Arabic text, helping you judge whether a passage suits learners or native readers. Runs entirely in your browser. It runs free in your browser on Gera Tools, with nothing uploaded.

Arabic Diacritic Density Meter

Name: Arabic Diacritic Density Meter
Creator: Gera Tools
License: https://creativecommons.org/licenses/by/4.0/

Get one useful tool a week

Like this tool? Enter your email and we'll send you one genuinely useful Gera tool a week — plus a link to come back to this one. No spam, one-click unsubscribe any time.

The Arabic Diacritic Density Meter reports what share of words in a passage carry tashkeel (harakat). Because fully vocalised Arabic is mainly used for learners, religious texts, and disambiguation, the density figure is a quick signal of how a text is intended to be read.

How it works

The text is split into words. For each word the tool checks whether it contains any Arabic diacritic codepoint:

fatha   ◌َ  U+064E     kasra   ◌ِ  U+0650
damma   ◌ُ  U+064F     sukun   ◌ْ  U+0652
shadda  ◌ّ  U+0651     tanwin  ◌ً ◌ٍ ◌ٌ  U+064B–U+064D
plus superscript alef U+0670 and related marks

A word is classed as vocalised if it carries at least one of these marks, and the density is the count of vocalised words divided by the total word count. The tool also reports the raw diacritic-mark count and average marks per word.

What different density ranges mean

Near 100 %: The text is fully or almost fully pointed. This is characteristic of the Quran, classical poetry editions, graded readers for language learners, and children’s books. Readers are not expected to supply vowels from context.

50–90 %: Selective pointing. Editors have marked words that are ambiguous, unusual, or frequently mispronounced, but the bulk of the text relies on reader knowledge. You find this in some academic editions of classical literature, in newspaper headlines where a new word is introduced, and in some educational materials aimed at intermediate learners.

Under 20 %: Effectively unvocalised. Standard Modern Arabic prose — news, novels, social media, business documents. Native adult readers expect to supply short vowels from context, morphological knowledge, and syntax.

Near 0 %: Completely unvocalised, typical of social media, informal messages, and OCR output from scanned printed material.

Practical uses for the density figure

Curriculum design. If you are preparing reading material for Arabic language learners, the density meter lets you confirm that the text has the level of pointing your students need. A passage intended for beginners should read near 90–100 %; a passage meant for advanced students should be well under 50 %.

Corpus analysis. When building an Arabic NLP training set, you may need to separate fully vocalised text (which is ideal for training diacritisation models) from unvocalised prose. Sorting by density is the fastest way to do this.

Typesetting quality control. Publishers adding harakat to classical texts use the density meter to catch sections where the editor forgot to vocalise a word, or where pasting text from another source introduced unvocalised fragments.

Reading-speed estimation. Fully vocalised Arabic takes longer to read aloud than unvocalised prose because every vowel is explicit. A very high density is a signal to use a slower reading-speed estimate.

Reading the average marks per word

A word can carry anywhere from one mark (a single sukun) to eight or more (a word with tanwin, shadda, and several harakat). The average marks per word, combined with the density percentage, gives a fuller picture. A text with 70 % vocalised words at 2 marks per word is lightly pointed; the same density at 5 marks per word is much more heavily marked.