A1Z26 Cipher Decoder and Encoder

Read each number as a count into the alphabet: 1 is A, 2 is B, and so on up to 26 for Z. So 3 15 4 5 reads C, O, D, E, which spells CODE. By hand it is quicker to work in fives, since 5 is E, 10 is J, 15 is O, 20 is T, and 25 is Y, which leaves at most four letters to count from the nearest landmark. An 18 is three past 15, so it lands on R.

To do it here, click the swap button and paste the numbers into the left panel.

A1Z26 is a name for the plain numbering of the English alphabet, and the name spells out its own rule: A is 1 and Z is 26. Every letter in between takes the number of its place in the alphabet, with no shift, no key, and no rearrangement, and the chart below gives the full mapping. Nothing needs memorizing beyond the alphabet you already know, which is why the cipher is so common in puzzles aimed at people who have never encoded anything before.

Reversing the cipher means turning the numbers back into letters, which is the same operation as decoding: take each number and count that far into the alphabet. Click the swap button to put this tool into that direction.

There is also a variant people call the reversed or backwards alphabet, where A is 26 and Z is 1. That is a different cipher, usually named Atbash. To convert between the two, subtract each number from 27: a 3 becomes 24, so it decodes to X rather than C. If a sequence produces nothing readable in standard A1Z26, trying it this way is a reasonable second guess.

No. A Caesar cipher shifts each letter a fixed number of places along the alphabet and gives you back letters, so with a shift of three, HELLO becomes KHOOR. A1Z26 does not shift anything; it swaps the alphabet for a different set of symbols, giving you numbers instead. HELLO becomes 8 5 12 12 15. The Caesar cipher tool handles the shifted kind, including working out the shift when you do not know it.

The two are often chained together in puzzles, with a Caesar shift applied first and the result then numbered, so if a decoded A1Z26 sequence produces letters that look like nonsense, a Caesar shift is the next thing to try.

ROT13 is a Caesar cipher with the shift fixed at 13, half the alphabet. Because 13 is exactly half of 26, applying it twice returns the original text, so the same operation both encodes and decodes, which is what the ROT13 tool does in a single direction.

A1Z26 has no such symmetry. Encoding turns letters into numbers and decoding turns numbers back into letters, so the two directions are genuinely different operations, which is why this tool needs a swap button and the ROT13 one does not.

The cipher has no official separator, which is the single thing that trips people up most. Without one the message is ambiguous: 1215 could be 12 and 15, which is LO, or 1, 2, 1, 5, which is ABAE. A separator between the numbers is therefore not decoration, it is what makes the message readable. Use the separator setting to pick a space, a dash, or a comma when encoding.

Decoding accepts all of those and a few more. Spaces, commas, hyphens, en and em dashes, slashes, pipes, and line breaks are all treated as separators and removed from the result. Line breaks are kept as line breaks, so a message laid out over several lines decodes with its layout intact.

A word break needs a wider gap than a letter break, or it disappears. When the numbers inside a word are joined by dashes or commas, a plain space is free to mean a word break, so 8-5-12-12-15 23-15-18-12-4 decodes as two words. When the numbers are separated by single spaces, the space is already doing the letter separating, and a word break has to be marked another way: a double space, a slash, or a line break all work.

That has a consequence when encoding. With the separator set to a space, a message of several words comes out as one continuous run of numbers and decodes back with no word breaks. Set the separator to a dash or a comma if the word boundaries matter to whoever is decoding it.

The cipher is case insensitive. Upper and lower case letters map to the same number, so A and a both become 1, and there is no way to record the difference in the numbers. Decoded output comes back in lower case for that reason.

Punctuation is not numbered. Anything that is not a letter passes through unchanged when encoding, so "hi!" becomes "8 9!" and the exclamation mark survives. Digits in the original text are the one awkward case: they pass straight through and are then indistinguishable from cipher numbers, so a message containing figures does not survive a round trip cleanly.

Nothing in the alphabet corresponds to it, so this tool leaves it in the output as a number rather than guessing. Seeing numbers above 26 usually means one of three things: the separators are missing or wrong and two numbers have run together, the message is not A1Z26 at all, or the numbers carry a second layer such as an offset that has to be removed first.

A zero is the same situation from the other end. There is no zeroth letter, so a 0 in a sequence is a sign the numbering starts from zero rather than one, in which case A is 0 and Z is 25, and adding one to every number before decoding will sort it out.

No, and it is worth being clear about why. A1Z26 has no key. Every other substitution cipher has something the recipient needs to know, even if it is only a shift of three; A1Z26 has nothing, because the mapping is fixed and published everywhere including this page. Anyone who recognizes the format can decode it. It is an encoding, in the same sense that Morse code is an encoding, rather than encryption.

What it provides is a moment of not being able to read something at a glance, which is what a puzzle wants and what protecting information does not. Never use it for anything that needs to stay private.

Look at the range. If every number falls between 1 and 26 and there is a clear separator between them, A1Z26 is the first thing to try. Frequency helps confirm it: in English the numbers 5, 20, 1, 15, and 9 turn up most often, because E, T, A, O, and I are the most common letters, so a sequence full of low and mid numbers with plenty of 5s is a good sign.

Numbers running up to 36 suggest letters and digits numbered together. Pairs of digits between 1 and 5, such as 11 24 33, suggest a Polybius square, where each pair is a row and column on a 5 by 5 grid. Numbers in the 65 to 122 range are more likely to be character codes than alphabet positions.

Almost entirely in puzzles. It shows up in escape rooms, where a decoded word gives the letters for a combination lock or the numbers give a padlock code directly, and in geocaching, where hints and coordinates are hidden in number strings on cache pages. Alternate reality games use it as an early step in a puzzle chain, usually as the layer that tells you the format of the puzzle rather than the layer that hides anything, and treasure hunts and puzzle books use it for the same reason.

Because it needs no key and no arithmetic beyond counting, it is also often the first cipher children meet, introducing the idea of a substitution before anything harder like a shift or a keyword.

The animated series Gravity Falls ends each episode with a short cryptogram in the credits, and across the first season those cryptograms rotate through three ciphers in turn: a Caesar shift of three, Atbash, and A1Z26. The show hides ciphers in signs, book pages, and background art as well, and the second season moves on to keyword ciphers that combine more than one method.

That rotation is why a large share of the people looking for an A1Z26 decoder are working on a Gravity Falls credit sequence. If the numbers decode to nonsense, the episode is probably on a different step of the cycle, and the Caesar cipher tool is the next thing to try.

Nobody in particular, and there is no useful history to give it. Numbering the letters of an alphabet is an obvious enough idea that it has been reinvented constantly, and versions long predate any name: gematria assigned numeric values to Hebrew letters, and the Polybius square of the second century BC numbered Greek letters by grid position. Neither is an ancestor of A1Z26 so much as other people arriving at the same idea. The name itself is modern and informal, from puzzle communities rather than cryptographic literature, where the construction is too trivial to have earned a name at all.