A1Z26 does not preserve capitalization. Every letter is reduced to its position in the 26-letter English alphabet — A is 1, B is 2, and so on up to Z at 26 — and that single numeric value carries no information about whether the original character was uppercase or lowercase. As a direct consequence, the input strings hello and HELLO both encode to the identical sequence 8-5-12-12-15, and any decoding step returns uppercase letters only. The reason is structural: A1Z26 has only 26 symbols available for 26 letter slots, so once a letter is assigned its number there is nowhere left in the format to record case. If a puzzle, game, or classroom exercise depends on knowing whether an original letter was capital, that information has to live in a separate channel — outside the numeric sequence itself. The A1Z26 Cipher Translator implements this rule exactly: it normalizes input to uppercase before encoding, and its decoded output is always uppercase A through Z. Anything else would require extending the cipher with extra conventions that the simple A=1 to Z=26 mapping was never designed to carry.
The fact that capitalization is lost in A1Z26 surprises many first-time users, because they expect letter substitutions to behave like ROT13 or the Caesar cipher, where each case is preserved independently. A1Z26 is not that kind of substitution. It is a position-numbering scheme that happens to be used as a cipher, and a position number is by definition case-neutral. Understanding this single distinction resolves most of the downstream questions about output formatting, equality of mixed-case inputs, and the impossibility of a "case-aware" A1Z26 variant without breaking the simple mapping.
Why A1Z26 Cannot Preserve Case
The mapping is built directly on the order of uppercase English letters in the ASCII basic Latin block. According to the Unicode Basic Latin chart, the code points for uppercase A through Z run consecutively, and the A1Z26 rule simply offsets that range so A maps to 1 and Z maps to 26. Nothing in that offset stores the original code point of the input character, so there is no place in the output to record whether the input started with an uppercase A (code point 65) or a lowercase a (code point 97). Both are reduced to the number 1, and once reduced the source code point is gone.
This is fundamentally different from case-preserving ciphers. ROT13, for example, applies a substitution that runs the alphabet forward by 13 positions independently for upper and lower case letters, so uppercase letters stay uppercase and lowercase letters stay lowercase. The Caesar cipher and Vigenere cipher behave the same way. Those ciphers keep two parallel alphabets — one for each case — and shift each along its own track. A1Z26 has no second track. It has a single number line of 26 values, and that line is case-neutral by construction. Adding case to A1Z26 would require doubling the output range, introducing a separate "upper" or "lower" marker, or pre-deciding an offset convention — every one of which violates the simple A=1 through Z=26 contract.
What Happens to Mixed-Case and Other Inputs
When the A1Z26 Cipher Translator receives mixed-case text, it normalizes everything to uppercase before computing the numeric sequence. This means that Hello World, HELLO WORLD, hElLo wOrLd, and any other case combination all produce the same output: 8-5-12-12-15 / 23-15-18-12-4. The slash in the middle is the tool's explicit word-boundary marker; hyphens separate letters inside a word, and the slash separates the two words. If you paste punctuation, digits, or accented letters like é, the encoder rejects the input with a specific error rather than silently dropping characters or inventing an escape sequence. The rejection is deliberate: an unsupported character would otherwise have to be either ignored (which loses information) or mapped to a numeric value (which invents a convention that the simple A=1 to Z=26 rule does not support).
| Cipher | Preserves Capitalization? | Underlying Reason |
|---|---|---|
| A1Z26 | No | 26 numeric slots for 26 letter positions; the value is case-neutral by definition |
| ROT13 | Yes | Two parallel alphabets (upper and lower) shift independently by 13 |
| Caesar | Yes | Each case is shifted along its own alphabet at the declared offset |
| Vigenere | Yes | A–Z key applied independently to upper and lower case letters |
The contrast in the table above is the cleanest way to internalize A1Z26's behavior. Case-preserving ciphers keep two parallel tracks of 26 letters; A1Z26 collapses both tracks into a single track of 26 numbers. Once that collapse happens, the original case cannot be recovered, because the information was never recorded in the first place. Treat A1Z26 as a one-channel system for letters alone.
How to Use the A1Z26 Cipher Translator
The translator is a small local tool with two modes — letters to numbers and numbers to letters — that runs entirely in the browser, so no input is uploaded. To answer the original question of whether A1Z26 preserves capitalization, the practical exercise is to encode the same word in two cases and observe the result.
- Open the A1Z26 Cipher Translator and select Letters to numbers.
- Type Hello in mixed case, then run the translator and copy the hyphen-separated numbers.
- Clear the input, type HELLO in all uppercase, run the translator again, and compare the output.
- Confirm both outputs are identical — 8-5-12-12-15 — which demonstrates in practice that A1Z26 does not preserve case.
- Add a second word such as World, separate it with a space, and run again; observe the slash that now appears between the two word groups (8-5-12-12-15 / 23-15-18-12-4).
- Switch to Numbers to letters, paste the hyphen-and-slash output, run, and confirm the decoded result is HELLO WORLD in uppercase regardless of the original case.
- Try a value outside 1–26 (for example 0 or 27) to see the explicit error the decoder raises, which confirms the strict range check.
These steps confirm the case behavior empirically. They also exercise the boundary marker: the slash separates words, hyphens separate letters, and an empty group between two slashes is rejected as an error. If you want to convert a phrase into a sequence where capitalization is meaningful — say, an acronym where the shape HTTPS matters — do not rely on A1Z26 alone. Treat the numeric sequence as a record of the letters only, and keep a separate note of which letters were capitals.
Tracking Capitalization When the Cipher Cannot
For puzzles where capitalization carries meaning, the safest workflow is to treat case as metadata, not as part of the cipher output. A few patterns work reliably across hand-written notes, classroom exercises, and lightweight puzzle sharing.
- Record a parallel pattern. For the word Hello, write the A1Z26 sequence 8-5-12-12-15 and a separate pattern such as 1-0-0-0-0 to mark the initial capital. The receiver applies the pattern after decoding to restore the case.
- Underline or mark capitals in plain text. Use a convention like H̲ello → 8-5-12-12-15 with the capital marked separately. This is the same idea as recording stress marks above letters in phonetics — a parallel channel that the cipher itself does not touch.
- Switch ciphers when case is part of the puzzle. ROT13, Caesar at a known shift, and Vigenere with a declared key all preserve case natively. If the puzzle needs to round-trip case, swap ciphers rather than try to bolt case onto A1Z26.
The choice is rarely "which cipher is better" and almost always "which property does the puzzle depend on." For word-length games, classroom substitution exercises, and lightweight number-puzzle tasks, A1Z26's loss of case is a feature: it makes the encoding simple, reversible by hand, and unambiguous for the receiver. For tasks where capitalization is part of the puzzle, the cipher is the wrong tool and the workaround above becomes a permanent extra rule rather than a one-off note.
Other Properties to Expect from A1Z26
Capitalization is not the only property that A1Z26 does not preserve. The mapping also discards punctuation, digits, and any character outside the ASCII A through Z range. Repeated whitespace is normalized to single word separators during encoding, so a pasted tab or a doubled space does not survive into the output. The cipher does preserve length (the number of letters in the input equals the number of hyphens plus one) and exact letter repetition (an input with two L's produces two 12's in the same position). It preserves no key, no randomized state, no integrity check, and no authentication. For a deeper round-trip walk-through that exercises both directions of the cipher on the same phrase, the A1Z26 Cipher Translate practical round-trip guide shows each step in detail and confirms the case behavior on a longer example.
The input limit of 200,000 Unicode code points is well above normal puzzle text and is there to keep the page responsive if an oversized paste is dropped in. Empty or whitespace-only input raises a clear error rather than silently producing an empty string. None of these constraints change the central answer to the original question: A1Z26 does not preserve capitalization. The format is defined by its 26 numeric slots, and any property that does not fit in those slots — uppercase versus lowercase, punctuation, digits, accented letters — is intentionally left out of the cipher and recorded separately if it matters to the puzzle being solved.
Related reading: Avoid Mistakes When Using an A1Z26 Cipher Translator.