Text Tools
Letter frequency analyzer.
Paste text to count how often each of the 26 letters appears, as counts and percentages, and to compute the index of coincidence, a statistic used in classical cryptanalysis. Accents are folded to the base letter. Your text stays in your browser.
What is counted
Only the 26 letters A to Z are counted, case-insensitively, with accented letters folded to their base letter. Digits, spaces and punctuation are ignored, and the percentages are shares of the counted letters. Short texts give noisy percentages, so treat a sample of a few dozen letters as a rough picture.
Index of coincidence
The index of coincidence (IC), from William F. Friedman's technique, measures how uneven the letter counts are. For N letters with counts ni in an alphabet of c = 26 letters, IC = c x sum of ni(ni - 1) divided by N(N - 1). Uniformly random letters give an expected IC of 1.0. According to the Wikipedia article, the actual monographic IC for telegraphic English text is around 1.73, and English plaintext will generally fall in the range 1.5 to 2.0 (normalised calculation). Some sources report the un-normalised value, which is the same divided by 26 (0.067 is 1.73/26, and 0.0385 is 1/26 for random letters).
Worked example
"Mary had a little lamb whose fleece was white as snow" has 43 letters, 17 different letters, most often "a" (6 times), and an IC of 1.641. The pangram "The quick brown fox jumps over the lazy dog" has 35 letters and an IC of 0.568, because it was written to use many different letters; very short texts can land far from the English range, so IC is only meaningful for longer samples.
Sources (each opened October 8, 2026): Wikipedia, Index of coincidence.
For shifting letters rather than counting them, see the ROT13 and Caesar cipher tool.
Letter frequency analyzer FAQ
What letters does the analyzer count?
Only the 26 letters A to Z, ignoring case. Accented letters are folded to their base letter, and digits, spaces and punctuation are ignored. Percentages are shares of the counted letters.
What is a normal index of coincidence for English?
The Wikipedia article on the index of coincidence says English plaintext will generally fall in the range 1.5 to 2.0 on the normalised scale, with about 1.73 for telegraphic English text, versus about 1.0 for uniformly random letters. Short texts vary a lot.
Is my text uploaded or saved?
No. The counting runs locally in your browser. This tool does not upload, store or submit the text you enter.