A becomes Z, B becomes Y. The same step encodes and decodes, so just type.
| A | B | C | D | E | F | G | H | I | J | K | L | M | N | O | P | Q | R | S | T | U | V | W | X | Y | Z |
| Z | Y | X | W | V | U | T | S | R | Q | P | O | N | M | L | K | J | I | H | G | F | E | D | C | B | A |
Write the alphabet forwards, then write it backwards underneath. Every letter swaps with the one opposite it: A↔Z, B↔Y, M↔N. The table above the explainer shows the whole thing. TREASURE becomes GIVZHFIV.
Because the swap goes both ways, Atbash undoes itself. Encoding and decoding are the same operation, and there's no key to find.
Atbash comes from Hebrew, where the first letter (aleph) swaps with the last (tav) and the second (bet) with the second-to-last (shin). Those pairs spell out the name: a-t-b-sh. Scholars have found it in the Book of Jeremiah, where Sheshach is read as Atbash for Babel.
It's a substitution, so word lengths and punctuation survive. Common words have distinctive Atbash forms: gsv is the, zmw is and. The cipher identifier tests for Atbash automatically. If Atbash comes close but not quite, the alphabet may have been reversed and then shifted. Try the substitution solver, which handles any one-to-one letter swap.
A substitution cipher that reverses the alphabet: A↔Z, B↔Y, C↔X and so on. It began with the Hebrew alphabet, and the name comes from its first and last letter pairs: aleph–tav, bet–shin.
Encode it again. Atbash is its own inverse, so running a message through twice gives back the original. That's why there's no encode/decode switch on this tool.
No. There's only one Atbash alphabet, so anyone who recognises it can read it straight away. Puzzle setters tend to use it as one layer among several.