Gematria

Every letter carries a number. Gematria adds them — a quiet mirror of the resonance in a name or word, not a fortune.

How is gematria calculated?

← Back to Bokki

How gematria is calculated

The hidden number in a name or word, using the letter-cipher that actually matches its script.

Hebrew text uses Mispar Hechrachi (the standard cipher) — א=1…י=10 … ק=100…ת=400, with final (sofit) letters taking their base value. Latin text uses A=1…Z=26, which we label plainly as a modern English extension, not the ancient tradition itself.

We never fabricate a Hebrew value from a guessed transliteration of a Latin name — Hebrew letters get the Hebrew cipher, everything else gets the English one. This is a mirror of numeric resonance, not a verdict.

How gematria is calculated

Is English gematria the same ancient tradition as Hebrew gematria?

No, and we say so on the page — English A=1…Z=26 is a modern extension inspired by the Hebrew and Greek tradition, not the tradition itself.

What if I type a name in Latin letters — do I get a Hebrew value?

No — we never guess a Hebrew transliteration for a Latin name. Hebrew script gets the Hebrew cipher, Latin script gets the English cipher, and we don't mix the two.

How many different totals this page can give

A sum does not care about order. Rearrange the letters of a name any way you like and the total does not move, which makes the row of letters this page prints a receipt you can check by eye rather than a path you could walk backwards from the total to the word. That has a consequence the method note does not mention: there are far fewer totals than there are things to type, and the gap is not a near thing. So we counted rather than described. Every possible string of letters at each length was pushed through the same function that answered you, with no sampling anywhere in it.

Every string of letters at each length, counted exactly rather than sampled. The Latin columns run over the twenty-six letters A–Z; the Hebrew columns over the twenty-seven characters alef through tav, with the final forms counted as spellings of their own. These are strings of letters and not a list of real names — we do not have one, and real names would sit differently inside the same range.
LettersLatin spellingsLatin totals reachedHebrew spellingsHebrew totals reached
126262722
267651729178
317,5767619,683577
4456,976101531,441976
511,881,37612614,348,9071,375
6308,915,776151387,420,4891,774
78,031,810,17617610,460,353,2032,173
8208,827,064,576201282,429,536,4812,572

On the Latin side the totals form an unbroken block. A word of five letters reaches every whole number from 5 to 130 and skips none of them — 126 totals standing behind 11,881,376 spellings. That is not luck at this one length: replacing any letter with the next one along adds exactly one to the sum, so the reachable set can never open a hole anywhere.

The Hebrew side is full of holes, and for a plain reason — those twenty-two values do not climb by ones the way A–Z does. They run from 1 up to 400, and the widest step between neighbouring values is 100 against a step of 1 the whole way along the Latin side, so sums land in clumps with empty ground between them. A single Hebrew letter can therefore be only 22 of the 400 whole numbers up to its own ceiling, and a two-letter word reaches 178 of 799. Whole stretches of the number line have no Hebrew word of that length behind them at all. The gaps do thin as words get longer — by eight letters the count is 2,572 — and they are still not closed at the longest text this page accepts.

There are five Hebrew characters that can never change a total. Alef through tav is twenty-seven characters on a keyboard but only twenty-two numbers in this cipher, because each final form keeps the value of the letter it is a form of: ך for כ (both 20), ם for מ (both 40), ן for נ (both 50), ף for פ (both 80) and ץ for צ (both 90). Write a word with a final mem or with an ordinary mem and the sum is identical. That is Mispar Hechrachi behaving as it should; the cipher that would separate those forms is a different one and is not in this code.

The box on this page holds forty characters, and that closes the answer space completely. In Latin script every total it can ever show is one of 1,040 — 1 through 1,040, with nothing missing in between. In Hebrew the ceiling is 16,000, and even with the whole box used not every whole number up to it is reachable: 15,379 of them are. The unreachable ones sit crowded near the top, where a total would need almost every character to be a tav, and the smallest of them is 15,311.

No total is a likely one. At eight letters the single busiest Latin sum is 108, and it holds 1.84% of the 208,827,064,576 spellings; at five letters the busiest share is 2.30%. Adding letters spreads the sum wider rather than piling it higher. A single letter has no peak at all — all twenty-six of its totals are equally common, because each of them has exactly one letter standing behind it.

What this page reads is the text in the box and nothing else. No birth date, no birth time, no gender, no place — there is one field on the screen and it takes characters. The language you are reading in changes the words around the number but never the number itself, because the cipher is chosen by the script you typed and not by the site's locale. And a shared total stays arithmetic. Two words adding to the same number is where the interpretive tradition begins, not a finding this page has made: with 11,881,376 spellings of five letters sharing 126 totals, agreement is what the counting does by default rather than a coincidence it has noticed.