Classical Cipher Decoder

Features

• Lists everything by default: ROT13, Atbash, ROT47, ROT18 and ROT5 at once, so you can spot the readable line
• Caesar takes any shift, and can brute force all 25
• Vigenère with your own key; the key advances only on key letters, so punctuation costs nothing
• Polybius square, both ways
• ⚠️ ROT47 is the only one that touches punctuation and digits, and the result explains why
• ⚠️ The square merges I and J, so decoding turns every J into an I — a property of the cipher, and stated in the result
• No Bacon cipher: it has two incompatible alphabets (24 and 26 letters) that encode to entirely different strings, and the text itself gives no clue which was used

How to use

1. Paste the text
2. The default “list all” shows every common transform; choose a specific method to narrow it down
3. Facing a Caesar cipher with an unknown shift? Turn on brute force and read down the list

FAQ

Why not shift every character?

Because classical ciphers are defined over letters only. Caesar, Atbash and Vigenère permute the alphabet; punctuation, digits and spaces take no part. Shifting those too destroys the readable structure and does not round-trip. The single exception is ROT47, which by definition rotates all of printable ASCII by 47.

Why are encoding and decoding the same operation for ROT13?

Because there are 26 letters and 13 is exactly half — rotate twice and you are back where you started. Atbash and ROT47 work the same way (the latter rotates 94 printable characters by 47, also half). So these need no direction setting, and this tool does not pretend they have one.

What happens to digits and symbols in a Vigenère key?

They are ignored; only letters are used. A key decides shift amounts, and a shift can only be 0–25, so digits and symbols have no meaning there. Note also that the key advances only on key letters: punctuation and spaces in the ciphertext do not consume a key letter, otherwise the same plaintext would fall out of step merely because it contained a few spaces.

Why is there no Bacon cipher?

Because it has two mutually incompatible alphabets: the original uses 24 letters (I/J and U/V share a slot) and the later version uses 26. The same run of A/B characters decodes to different text under each, and nothing in the text reveals which was used. Handing over a “decoded” answer the user cannot check — one that may be half wrong — is worse than not offering it.