How to count words and characters in Arabic text
The Counter gives you words, characters, lines, paragraphs and sentences, plus an estimated reading time. You need it when an essay or article has a word limit, when you price a translation per word, when a bio or ad field caps the number of characters, or when you want to know how long a speech or video script will take to read.
Arabic counts depend on the convention you use. Is a vowel mark a character? Is an emoji one character? Is والكتاب one word or two? So you pick the convention, and the report records it next to the numbers.
How to use it
- Paste your text into Text to count. The numbers appear straight away.
- Under Words, choose Linguistic words (browser) for a count that skips punctuation and stray symbols, or Whitespace-separated tokens to count everything between two spaces.
- Under Characters, choose a unit: Graphemes / complete emoji, Unicode code points or UTF-16 code units. The difference is explained below.
- Decide on Include whitespace and line breaks and Include numeric tokens as words, depending on what the person receiving the text expects. Both start switched on.
- Set Reading speed: words/minute anywhere from 50 to 1,000. It starts at 200.
- Copy or download the Count report. It lists the numbers together with the settings you chose.
Worked examples from the tool
مَرْحَبًا 👋🏽
| Words | 1 |
|---|---|
| Characters | 7 |
| Lines | 1 |
| Non-empty lines | 1 |
| Paragraphs | 1 |
| Sentences | 1 |
| Reading time (seconds) | 1 |
Settings: Characters: Graphemes / complete emoji
مَرْحَبًا 👋🏽
| Words | 1 |
|---|---|
| Characters | 12 |
| Lines | 1 |
| Non-empty lines | 1 |
| Paragraphs | 1 |
| Sentences | 1 |
| Reading time (seconds) | 1 |
Settings: Characters: Unicode code points
تبدأ المسابقة يوم الأحد. سجّل قبل الموعد. تُعلن النتائج بعد أسبوع؟ نعم، على الموقع.
| Words | 14 |
|---|---|
| Characters | 82 |
| Lines | 3 |
| Non-empty lines | 2 |
| Paragraphs | 2 |
| Sentences | 4 |
| Reading time (seconds) | 5 |
Settings: Default settings · 200 words per minute
Which character unit should you use?
- Graphemes / complete emoji count what you see as one character. A letter with its vowel marks is one unit, and so is an emoji with a skin tone. Use it for most limits written for people, such as “no more than 300 characters”.
- Unicode code points: a code point is the number Unicode assigns to each character. Here every vowel mark counts separately, so vowelled text gives a higher number.
- UTF-16 code units are how JavaScript and some other systems measure text length. An Arabic letter is one unit; many emoji take two or more. Choose it when you know the destination system counts this way.
In the two examples below, a vowelled word plus an emoji gives 7 graphemes and 12 code points, and it reaches 14 in UTF-16 units.
Why word counts differ between tools
The tool does not split off attached conjunctions and prepositions, so والكتاب and بالقلم are one word each in both modes. A dash or symbol standing between two spaces counts as a word in Whitespace-separated tokens but not in Linguistic words (browser). Numbers such as 1250 count as words unless you turn off Include numeric tokens as words.
A paragraph, for the tool, is a block separated by a blank line. Text copied from a PDF often arrives with every printed line on its own line, which inflates the line count. Join the lines first in the Cleaner, as described in cleaning up copied Arabic text.
Word frequency
Below the numbers you will find the Word frequency table: each word, how often it appears and its share of the total. Tick Ignore common marks in frequency and Normalize alef in frequency so that أحمد and احمد land in the same row, and list words such as في من على under Your stopwords, separated by whitespace. The table shows the first 200 words; Export all frequencies CSV includes them all.
Limits
- Up to 100,000 text units.
- Word counts are estimates based on your browser's text segmentation and can vary slightly between browsers and versions.
- There is no morphological analysis, so attached و, ب and ل are never split off.
- Sentence counts depend on punctuation and line breaks; a long line with no punctuation counts as one sentence.
- Reading time is simple arithmetic: words divided by your chosen speed, rounded up to the next second. It does not measure how you actually read.
- Numbers can differ from other programs such as Word, because each program has its own counting rules.
Frequently asked questions
Do vowel marks count as characters?
Not in Graphemes / complete emoji, where a letter and its marks are one unit. In Unicode code points and UTF-16 code units, every mark counts separately.
Is والكتاب one word or two?
One, in both word modes. The tool does not split attached و, ب or ل from the word.
Why is my count different from Word or Google Docs?
Each program decides in its own way what counts as a word or a character, including dashes, numbers and symbols. Pick the convention closest to what the recipient expects and state it if precision matters. The report writes it down for you.
How do I estimate how long a speech or voice-over will take?
Read a familiar passage aloud for one minute, count its words here, and enter that number as Reading speed: words/minute. The tool then times your new text at the same pace.
Does an emoji count as one character?
Yes, in Graphemes / complete emoji in modern browsers, even when it is built from several parts such as a skin tone. In the other two units it usually takes more than one.