How does the Hill cipher encrypt text?

Letters are mapped to numbers A=0 to Z=25 and grouped into pairs (digraphs). Each pair is treated as a column vector and multiplied by the 2×2 key matrix modulo 26, producing a new pair of numbers that map back to ciphertext letters.

When is a Hill key matrix invalid?

The key works only if its determinant is coprime with 26 (the determinant's greatest common divisor with 26 must be 1). If the determinant shares a factor of 2 or 13 with 26, no modular inverse exists and the message cannot be decrypted, so the tool rejects such keys.

How is the decryption matrix calculated?

Decryption uses the modular inverse of the key matrix mod 26. For a 2×2 matrix this is the modular inverse of the determinant multiplied by the adjugate (swap a and d, negate b and c). The result undoes the encryption multiplication.

Why is the message padded with X?

Because the cipher works on pairs of letters, a message with an odd number of letters has one leftover letter. A filler letter, conventionally X, is appended so the final pair is complete. You can strip a trailing X after decryption if it was padding.

Is the Hill cipher secure?

No. It is linear, so it falls quickly to a known-plaintext attack: a few known digraph pairs let an attacker solve for the key matrix. It is valuable as a teaching example of linear algebra in cryptography, not for real security.

What is the Hill Cipher?

Free Hill cipher encoder and decoder using a 2×2 key matrix. Encrypts digraphs by matrix multiplication modulo 26 and decrypts with the modular matrix inverse, checking that the key determinant is invertible mod 26. Shows the inverse key. Runs entirely in your browser. It runs free in your browser on Gera Tools, with nothing uploaded.

Hill Cipher — Gera Tools

Name: Hill Cipher
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 Hill cipher, invented by Lester S. Hill in 1929, was the first practical cipher to encrypt more than one letter at a time using linear algebra. Instead of substituting single letters, it treats blocks of letters as vectors and multiplies them by a key matrix modulo 26. This tool implements the classic 2×2 version, which encrypts the message two letters at a time.

How it works

Each letter is converted to a number with A = 0, B = 1, … Z = 25. The plaintext is split into pairs of letters, and each pair is written as a column vector. To encrypt, the vector is multiplied by the 2×2 key matrix and every result is reduced modulo 26:

[ a b ] [ p1 ]   [ (a·p1 + b·p2) mod 26 ]
[ c d ] [ p2 ] = [ (c·p1 + d·p2) mod 26 ]

The two output numbers map back to ciphertext letters. To decrypt, the ciphertext vector is multiplied by the modular inverse of the key matrix. For a 2×2 matrix the inverse is the modular inverse of the determinant det = (a·d − b·c) mod 26 multiplied by the adjugate matrix [d, −b; −c, a]. This only exists when det is coprime with 26, so the tool checks that condition before encrypting.

Worked example

With key matrix [3 3; 2 5] the determinant is 3·5 − 3·2 = 9, and gcd(9, 26) = 1, so the key is valid. The plaintext pair HE becomes numbers (7, 4). Multiplying gives (3·7 + 3·4, 2·7 + 5·4) = (33, 34), which reduces mod 26 to (7, 8) — the ciphertext letters HI. Decryption multiplies by the inverse key to return (7, 4) and recover HE.

Notes and tips

Choose a key whose determinant is coprime with 26 — avoid even determinants and multiples of 13. Messages are upper-cased and non-letters are removed before encryption, and an X is appended if the letter count is odd so the last pair is complete. The Hill cipher is linear and therefore breakable with known plaintext; use it for learning, not for protecting secrets. All computation runs locally in your browser.