Skip to content
TextLab

Free letter counter with a frequency table

Letter counter with every letter broken out

TextLab's letter counter counts letters and nothing else — digits, spaces and punctuation stay out of the total and are reported on the side. It is free, no account is involved, and the totals move on every keystroke. Under the total sits a row for each of the 26 letters with its count, its share of your text and the figure it should be near in ordinary English, plus a vowel-to-consonant split and a grid that tells you when you have written a pangram.

  • 100% free
  • No signup
  • Digits and spaces excluded
  • All 26 letters counted
  • Pangram check
0lettersout of 0 characters · 0 vowels, 0 consonants

0
Letters
0
Vowels
0
Consonants (with y)
0
Letters of 26 used
0
Characters in total
0
Digits left out
0
Spaces left out
0
Punctuation left out

Each letter, and how often

  • a
    00.0%8.2%
  • b
    00.0%1.3%
  • c
    00.0%2.8%
  • d
    00.0%4.3%
  • e
    00.0%12.7%
  • f
    00.0%2.2%
  • g
    00.0%2.0%
  • h
    00.0%6.1%
  • i
    00.0%7.0%
  • j
    00.0%0.1%
  • k
    00.0%0.8%
  • l
    00.0%4.0%
  • m
    00.0%2.4%
  • n
    00.0%6.8%
  • o
    00.0%7.5%
  • p
    00.0%1.9%
  • q
    00.0%0.1%
  • r
    00.0%6.0%
  • s
    00.0%6.3%
  • t
    00.0%9.1%
  • u
    00.0%2.8%
  • v
    00.0%1.0%
  • w
    00.0%2.4%
  • x
    00.0%0.1%
  • y
    00.0%2.0%
  • z
    00.0%0.1%

The last column is how often that letter appears in ordinary English. A big gap between the two — a page where q or z is doing far more work than 0.1% — usually means names, jargon or another language mixed in.

Alphabet coverage

abcdefghijklmnopqrstuvwxyz

A pangram uses all 26 letters at least once. Type one — the quick brown fox jumps over the lazy dog — and every square above fills in.

Vowels against consonants

English runs at roughly 38 vowels for every 62 consonants; the bar shows where your text sits against that.

How to count the letters in a text

Three steps, and the alphabet grid starts filling from the first word.

  1. Put the text in and watch the digits and spaces drop out

    Type or paste anything — a name, a headline, a puzzle answer, a whole chapter — or hit the paste shortcut wherever your cursor happens to be: ⌘V on a Mac, Ctrl+V on Windows. The headline figure counts letters alone: the 1s and 7s in a phone number, the spaces between words and every comma and full stop are excluded and reported separately so you can see what was left out.

  2. Read the table of all 26 letters

    Every letter of the alphabet gets a row with its count, its share of your letters and, in the last column, how often that letter turns up in ordinary English. Switch the order to “Most used first” to see your own top letters, and compare the two percentage columns to spot a page where names or jargon have pushed a rare letter far above its usual 0.1%.

  3. Check the alphabet grid and the two switches

    The grid of 26 squares fills in as each letter appears, so a pangram lights up completely and anything short of one names the letters still missing. Use “Count é, ü and å with their plain letters” to fold accented letters into their base letter or keep them separate, and “Treat y as a vowel” to move y between the vowel and consonant totals.

Technical specifications

What counts as a letterAny Unicode letter — the 26 English letters in either case, plus accented and non-Latin letters. Digits, spaces, tabs, line breaks, punctuation, symbols and emoji are all excluded and counted separately
Frequency tableAll 26 letters with their count, their share of your letters and the English average beside it, sortable A to Z or most-used first
Reference frequenciesRobert Lewand's table from Cryptological Mathematics: e 12.70%, t 9.06%, a 8.17%, o 7.51%, i 6.97%, n 6.75%, down to q 0.10% and z 0.07%
Accented lettersFolded into their base letter by default, so é joins e and ñ joins n; turn folding off and they are listed separately as letters from outside the English alphabet
Vowels and consonantsa, e, i, o and u are vowels; y is a consonant unless you switch it, which moves roughly 2% of a typical English text from one side to the other. Ordinary English runs near 38% vowels
Alphabet coverageA 26-square grid that fills as letters appear, naming the letters still missing and confirming a pangram once all 26 are present
Length limit and speedNo cap. A 5,000-character cover letter is counted in 0.038 ms and a 600 KB manuscript in 3.6 ms on an Apple M2 Pro laptop, in one pass over the text
PriceFree, with no signup, no daily limit and nothing to install

Frequently asked questions

Do spaces count as letters?

No — a space is not a letter, and neither is a digit, a comma or a dash. That is the whole difference between this page and a character count: “Hello, world 2026!” is 18 characters but 10 letters. The four tiles under the box show exactly what was excluded — digits, spaces, punctuation and symbols — so the numbers always add up and you can see where the difference went.

Does é count as a letter, and as which one?

It always counts as a letter; whether it counts as an e is a switch you control. With folding on, which is the default, é, è and ê all add to the e row, ñ adds to n and å adds to a, which is what you want when you are analysing English text peppered with café, naïve and résumé. Switch folding off and they leave the 26 rows entirely and are reported as letters from outside the English alphabet, listed individually. German ß has no single base letter and always sits outside; Greek, Cyrillic and Chinese characters are counted as letters in the total but have no row of their own.

Is y a vowel or a consonant here?

A consonant by default, with a switch to move it. English has no clean answer: y is a vowel in rhythm, myth and happy, a consonant in yellow and yet, and both in yolky. It appears in about 2 per cent of English letters, so the choice moves the vowel share by roughly two points — enough to matter if you are comparing two texts, which is why the setting is visible rather than buried.

What is a pangram, and how do I know if I have written one?

A pangram is a sentence that uses all 26 letters at least once, and the grid above turns solid when yours does. The famous one — the quick brown fox jumps over the lazy dog — spends 35 letters doing it. A perfect pangram uses exactly 26, one of each, and English has very few that make sense without proper nouns or obscure words. Until the grid is full, the missing letters are named underneath it, which is faster than checking j, q, x and z by eye every time.

Which letter is most common in English?

E, at about 12.7 per cent of all letters, followed by t, a, o, i and n. The last column of the table above carries the full set of figures from Robert Lewand's Cryptological Mathematics, the count cryptographers have used for decades. They hold up remarkably well on any ordinary English page of a few hundred words, which is what makes them useful: a text where your own percentages wander far from them is usually full of names, technical terms, another language, or is simply too short for the averages to mean anything.

Why do I need a letter count rather than a character count?

Because some limits and puzzles are about letters specifically, not everything you typed. Crossword and Wordle answers, Scrabble racks, monogram and initials work, cryptogram solving, alliteration checks in a headline, and any assignment phrased as “write a 500-letter summary” all mean letters alone. Anything counted by a web form, a text message or an X post means characters, spaces and punctuation included — that is a different number, and the character counter is the page for it.

Does it count capitals separately from lower case?

No, A and a share a row, because the question people are asking is how often the letter appears rather than how it was typed. That is also how letter-frequency tables, crossword grids and Scrabble tiles treat the alphabet. If the split between capitals and lower case is what you are after, a case converter shows it directly: convert a copy of the text to lower case and compare the two letter counts.

Twenty-six letters, and the shape they usually make

English letters arrive in remarkably stable proportions. Take any few hundred words of ordinary prose and roughly one letter in eight is an e; t, a, o, i and n follow in a predictable order; and j, q, x and z between them account for less than half a per cent of the page. The classic figures — e at 12.70%, t at 9.06%, a at 8.17%, down to z at 0.07% — come from Robert Lewand's count in Cryptological Mathematics, and they are the reason a simple substitution cipher can be broken with a pencil: the letter that appears most in the ciphertext is almost certainly standing in for e. The same stability is why the last column of the table above is worth reading. When your own percentages sit close to it, your text is behaving like English. When they do not — z at three per cent, e down at seven — the cause is usually a name repeated forty times, a technical term, a second language mixed in, or simply a text too short for averages to mean anything at all.

The letters-only total exists because plenty of tasks care about letters and nothing else. Crossword answers and Wordle guesses are counted in letters; a Scrabble rack is seven tiles, not seven characters; a monogram is three letters whatever the spaces around it do; a cryptogram is solved by frequency; and alliteration in a headline is a question about first letters. Meanwhile a lot of the world counts differently. A web form, a text message, a job portal or a database column counts characters, so the space after every word and each comma and full stop takes up room — the character counter is the page for that, and it is often a fifth higher than the letter figure on the same text. Between the two sits the ordinary word count, which the word counter keeps alongside characters, sentences and paragraphs.

Two decisions in any letter counter are arguable, so both are switches here rather than silent defaults. The first is what to do with an accent. Fold é into e and a page of French borrowings reads as English, which is what you want when comparing frequencies; leave it separate and you can see exactly how much of the text is not plain a-to-z. The second is y. It works as a vowel in rhythm and happy and as a consonant in yellow, it is about 2 per cent of English letters, and there is no answer that is right for every text — so the total moves when you decide, rather than the page deciding for you. Once you know which letters dominate a draft, the word frequency counter takes the same question one level up and shows which whole words you keep reaching for.

Nothing leaves this tab

The letters are tallied by code that came down with the page, so the puzzle answer, client headline or private list you paste stays on your own computer — there is no upload step and no request to a server anywhere in the process. The two switches are not remembered between visits and neither is your text: reloading gives you an empty box and the default settings back.