What the Atbash cipher does
The Atbash cipher is one of the oldest substitution systems in written history. Instead of shifting letters forward like Caesar or ROT-style ciphers, Atbash mirrors a character against the end of its alphabet. In the English alphabet that means A becomes Z, B becomes Y, and so on. In Hebrew, the same reversal principle creates the original historical form of Atbash.
This Atbash cipher tool lets you work with English Latin text, Hebrew text, and even a fully custom alphabet that you define yourself. Because Atbash is its own inverse, the same transform works for both encoding and decoding. Apply it once to scramble the text, and apply it again to recover the source.
Why people use Atbash
Atbash is useful for educational demos, puzzle design, escape-room clues, lightweight obfuscation, and language learning. It is also a good way to explain how monoalphabetic substitution works because the rule is easy to verify visually from a mapping table.
This is not modern encryption. Anyone who knows the alphabet or sees a few examples can reverse it. Use it for curiosity, classic cryptography exercises, or playful text transformations, not for secrecy.
What makes this tool practical
The interface gives you real-time output, a full alphabet mapping table, grouped 5-character ciphertext formatting, clipboard copy, downloads, and shareable links. You can keep punctuation intact, strip it out, or replace it with spaces depending on how traditional or readable you want the output to be.
Everything runs client-side in the browser, so nothing is uploaded when you transform text. That makes the tool fast, private, and easy to use on both desktop and mobile.
Latin, Hebrew, and custom alphabet support
For Latin text, the tool supports case-preserving output plus forced uppercase or forced lowercase modes. That makes it useful both for readable notes and for traditional all-caps ciphertext styling. If you switch to grouped output, the tool strips everything outside the active alphabet and presents the result in classic 5-character blocks.
Hebrew mode follows the original Atbash idea with the standard 22-letter order. Final letters are normalized for mapping and can be restored at word endings so the output still reads naturally. This matters for letters such as ך, ם, ן, ף, and ץ.
Custom alphabet mode lets you define any ordered set of unique characters. That means you can build mirrored transforms for uppercase letters only, lowercase letters only, mixed symbol alphabets, classroom exercises, or puzzle-specific character sets. The only rule is that each character must appear once.
Options that change the result
Atbash output depends on more than the alphabet itself. Non-letter handling controls whether punctuation, emojis, and symbols stay in place, disappear, or become spaces. Separate digit handling lets you either keep numbers untouched or mirror them as an optional 0↔9, 1↔8, 2↔7 pattern. Whitespace normalization is helpful if you want neat results from pasted paragraphs or copied chat logs.
The statistics panel helps you understand what changed. You can see how many letters, digits, symbols, and spaces were present in the source and what share of characters were actively transformed. That is especially helpful when you are testing custom alphabets or using grouped output that intentionally removes content.
Example transformations
A classic English example is HELLO WORLD becoming SVOOL DLIOW. If you run the transform a second time, the original text comes back immediately. Hebrew works the same way, but with its own letter order and optional final-form handling.
If you need a playful demo, define a custom alphabet like abc123. The first character mirrors the last, the second mirrors the second-to-last, and the center stays fixed when the alphabet length is odd. That makes custom mode useful for puzzle authors, teachers, and anyone experimenting with reversible substitution rules.